THE UNIVERSITY OF OKLAHOMA GRADUATE COLLEGE

PARMAMETRIC COST ESTIMATING MODEL, TO ESTIMATE ENGINEERING DESIGN FEES, FOR MULTI/ TECH ENGINEERING SERVICES, SALEM, OREGON

A SPECIAL RESEARCH PROJECT SUBMITTED TO THE DIVISION OF CONSTRUCTION SCIENCE IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE DEGREE OF MASTERS IN CONSTRUCTION ADMINISTRATION

BY

ASHEESH BAJAJ FALL 2001

Acknowledgements Achieving is locating a goal and then finding the best path to reach it, and higher the summit, the harder the climb. After descending upon the goal, I have worked from part to whole, dividing the total scope of work into fragments and working on them one by one. Of course in the end, bringing them all together. It has been rightly said that we are built on the shoulders of others. This dissertation has been made possible with the guidance, support, keen insight, endless advice, incredible patience, as well as the more concrete organization, and direction, of my committee chair Professor Doug Gransberg. My very special thanks to Mr. Mark Grenz, President, Multi/Tech Engineering Services, for providing the precocious cost data from his firms accounting history, and also funding the research. I would like to thank Jessica, Amy, and Dan from Multi/ Tech for responding to all queries, and information requests. I would also like to thank Mr. Bob Fillpot, Dean, School of Architecture, and Mr. Thomas Hardy, in charge of the computer labs, of School of Architecture for providing a laptop for the summer, for research purposes. Finally, I would also like to thank my wife, my brother, and my parents for their constant support and encouragement.

CONTENTS Introduction Need For The Study The Case Project Objectives Scope Of Work Literature Review ------4 4 4 5 6 7 7 7 8 9 -14 14 16 17 18 20 20 22 22 24 24 25 27 28 28 28 30 33 33 33 35 36 36

Statistics -Measure Of Central Tendency -Measures Of Dispersion -Scatter Plot -Confidence Level 12 Outliers In Regression -Grubbs Method For Assessing Outliers -Allen Blumans Method For Detecting Outliers -Wilfrid Dixon, And Frank Massys Method Of Finding Outliers. Research & Education Associations Method Of Finding Outliers Parametric Estimation -Construction Industry -Design Industry -Effective Cost Accounting System -Prepare Database From The Accounting History. -Identify The Different Multipliers Required To Normalize -The Data Exclude Outliers -Identify The Important Dependent Variables -Do Statistical Analysis To Find Cost Estimating Relationships. Apply Required Multipliers. -Prepare A Template -Methodology Estimation Of Design Fee - Multi Tech Engineering Services Data Collection Preparing The Database Tread Line Graphs Dealing With Outliers Analysis --------

Testing the Parametric Cost Estimating Model Model For All Projects Model For Apartment Projects Model For Subdivision Projects Model For Miscellaneous Projects Conclusions And Recommendations Limitations of the Database Areas for Future Research Bibliography

----------

45 41 42 43 44 53 53 54 55 57 63 72 80 80

Appendix I: Definitions Of Terms -Appendix II: All Relationships Without Excluding Outliers -Appendix III: All Relationships After Excluding Outliers -Tables Table A ---

Introduction Design cost is a very important aspect for a project. This importance is not evident in the beginning, but is clearly seen after the final completion of the project. The design cost is directly reflected by the quality of design for a project. The designer must apportion design efforts to fit available funding, thus, underestimating the design fee creates an economic bias to minimize design product. Need For The Study The design cost estimation scenario, in most of the design firms, including, architectural, engineering (civil, structural, mechanical, and electrical) and urban planning firms, is not as clear-cut, as construction cost estimation scenario. As a result of this, the implications of the design cost, if it is estimated lower that what it should have been, are reflected in design after the project is completed. It is very important to accurately estimate the design costs, to achieve quality design. The design costs can be estimated on the basis of existing data from a design firm. The analysis of this data would help in preparing a template to estimate design costs for that particular firm. This template may vary from one firm to another. The project at hand, will primarily focus on data gathered from Multi Tech Engineering Services, an engineering design firm, in Salem, Oregon. The study will conclude with the preparation of a design fee estimating template, for Multi Tech Engineering Services. The Case Multi/ Tech Engineering Services is an engineering design firm, which primarily does subdivision projects in Oregon. The subdivision includes, laying out the plots, and the road network, and proposal of services. Land survey of the site is also an inherent part of their work. They have enough survey work to sustain a surveying team of their own. Their work also includes design and construction management of the apartment complexes, built on these subdivisions. Multi/ Tech Engineering Services has complete data from their past 200 projects. This data includes the exact number of man-hours spent in design and construction management on each project. The man-hours are further divided into technician (AutoCAD drafts person) man-hours, Surveyor man-hours, Engineer man-hours, Supervisor man-hours, and Construction Manager man-hours. The data includes figures on the total estimated man-hours at each level, and the total actual utilized man-hours at each level. It also includes total man-hours paid. The primary aim of this research is to organize this data into a database, conduct statistical analysis, prepare a template for estimating the total number of man-hours required, and in turn design fee for each type of project.

Project Objective The objectives of the research are as follows: Organization of the existing data into fields. Statistical analysis of the organized data. Identification of all aspects affecting the design costs, for the specific project types done by the Multi- Tech Engineering Services. Analysis of identified aspects affecting the design costs of a project. Preparation of a template to estimate the design fee for Multi-Tech Engineering Design Firm. Test the template with several projects not included in the database.

Scope of Work The scope of work for this study will be limited to the to organizing, and doing analysis on the data. Never the less a calculator will be proposed, but the final decision of the design fee will be left on the personnel from Multi/ Tech, as there are many other factors affecting the design fee, that they understand better than the author of this study.

Literature Review The literature review primarily includes the study of Statistics, and Parametric Estimation to an extent, such that after reading the literature review, one can understand the concepts of Statistics, and Parametric Estimation used in this study. Statistics It is important to understand concepts of statistics, to effectively use statistical analysis for parametric estimating. The concepts that were used in this research are (Research and Education Foundation, 1996): Measure Of Central Tendency Measures Of Dispersion Scatter Plot Confidence Level

Measure of Central Tendency The measures of central tendency are Mean, Median, and Mode. (Research and Education Foundation, 1996) Mean: is one of the measure of central tendency. It is calculated by adding all the data points, and dividing this figure by the population size. Following is an example for calculating mean. Calculate the mean unit cost for division 9 from the unit cost of division 9 from eight different projects. ($10, $10.5, $11, $9, $19, $7,$10, & $9). 10 + 10.5 + 11 + 9 + 19 + 7 + 10 +9 Mean = 8 = $10.69 Median: is also one of the measures of central tendency in the data. It is the halfway point in a data set. Before finding this point, the data must be arranged in ascending or descending order. A data set in ascending or descending order is called a data array. The median will either be a specific value in the data set, or will fall between two values, as shown in the following example. Calculate the median of unit cost for division 9 from eight different projects. ($10, $10.5, $11, $9, $19, $7,$10, & $9). Arrange the data in order: 7, 9, 9, 10, 10, 10.5, 11, 19

10

The median in this case is :

10 + 10 = 10 2 If this data set was comprised of seven data points instead of eight, then the 4th data point would have been the median. Mode: is also one of the measures of central tendency in the data. It is the value that occurs most often in the data set. It is sometimes also referred to as the most typical case. Calculate the mode for unit cost for division 9 from seven different projects. ($10, $10.5, $11, $19, $7,$10, & $9). Arrange the data in order: 7, 9, 10, 10, 10.5, 11, 19 Mode = 10

Measures of Dispersion The measures of dispersion are Variance, Standard Deviation, and Range. (Bluman, 1998) Variance: is the average of squares of distance, each value is from the mean. The symbol for variance is 2 The formula for the population variance is: (X )2 2 = N Where X = Individual value = Population mean N = Population size Variance for the data set ($10, $10.5, $11, $9, $19, $7,$10, & $9) will be: (10-10.69) 2 + (10.5-10.69) 2 + (11-10.69) 2 + (9-10.69) 2 + (19-10.69) 2 + (7-10.69) 2 + (10-10.69) 2 + (9-10.69) 2} 8 0.692 + 0.192 + 0.312 + (-1.69)2 + 8.312 + (-3.69)2 + 0.692 +(1.69) 2 8 = 12.78 =

2 =

11

Standard Deviation: is the square root of the average of, squares of distance each value is from the mean. The symbol for variance is (X )2 N Where X = Individual value = Population mean N = Population size Standard Deviation for the data set ($10, $10.5, $11, $9, $19, $7,$10, & $9) will be: (10-10.69) 2 + (10.5-10.69) 2 + (11-10.69) 2 + (9-10.69) 2 + (19-10.69) 2 + (7-10.69) 2 + (10-10.69) 2 + (9-10.69) 2} = 8 = 0.692 + 0.192 + 0.312 + (-1.69)2 + 8.312 + (-3.69)2 + 0.692 +(1.69) 2 8

= 12.78 = 3.57 Range: for a data set is the lowest value subtracted from the highest value in the data set. It is represented by R. Range for the data set ($10, $10.5, $11, $9, $19, $7,$10, & $9) will be: R = 19 7 = 12 Scatter plot Scatter plot is the plot of all the data points of one data set, with the corresponding data points of another data set (Research and Education Foundation, 1996). The Dependent variable is plotted on the y axis, and the independent variable is plotted on the x axis. An example of the scatter plot is shown in Figure 1.

12

S.No. 1 2 3 4 5 6 7 8 9

Cost of projects vs. the unit cost of the division 1 for each project. Project Cost $142,774.00 $140,253.00 $134,671.00 $125,934.00 $116,028.00 $277,151.00 $352,945.00 $247,400.00 $222,089.00 Division 1 cost 16.58 14.21 13.74 16.14 21.12 11.53 12.00 5.85 7.51

Project Costs Vs Unit Cost of Div 1 $400,000.00 $350,000.00 $300,000.00 $250,000.00 $200,000.00 $150,000.00 $100,000.00 $50,000.00 $0.00 0.00 5.00 10.00 15.00 20.00 Unit Cost Of Div 1

Project Cost

13

Figure 1: Scatter Plot Trend Line: If we join the points in a scatter plot, a line can be formed. This line is called a trend line. The data points do not always fall in the same line, and joining them can make a zigzag pattern sometimes. This is because of outliers in the data, which can distort the trend line. Outliers will be addressed later in this study. A trend line minimizes the squared distances, between the y values that lie on the chosen line, and the observed y values. It can be plotted with the help of Microsoft Excel. Plotting a trend line manually is out of the scope of this study. Trend lines can be Linear, Polynomial, Logarithmic, and Exponential. Linear trend line for the scatter plot in Figure 1, plotted with the help of MS Excel is shown in Figure 2.

Chart Title $400,000.00 $350,000.00 $300,000.00 $250,000.00 $200,000.00 $150,000.00 $100,000.00 $50,000.00 $0.00 0.00 5.00 10.00 15.00 20.00 Unit Cost Of Div 1

Figure 2: Linear trend line for a scatter plot

Project Cost

14

Confidence level Confidence level in regression can be judged with the help of Standard Error, R Squared Value, Coefficient of Variance, and Skewness (Research and Education Foundation, 1996). Standard Error: The mean squared deviation of sample points from the estimated regression line. It is a measure of the amount of error in the prediction of y for an individual x. It can be calculated with the following equation:

1 Sy.x = n(n 1) Where n is the population size. x is the independent variable y is the dependent variable ny2 (y)2 --

[nxy - (x)(y)]2 nx2 -- (x)2

R Squared Value represents the degree of accuracy of dependent variable calculated with the help of an independent variable and an equation. This value ranges from 0 1.0. Examples of R-squared value can be seen in Figures 3 and 4.

15

Chart Title $500,000.00 Project Cost $400,000.00 $300,000.00 $200,000.00 $100,000.00 $0.00 0.00 5.00 10.00 15.00

y = 16044x + 32114 R2 = 0.9639

20.00

25.00

30.00

Unit Cost Of Div 1

Figure 3: R Squared Value In Figure 3, since the plotted points clearly define the trend line, the R-squared value is near to 1.0, which means that 96% of the variance in the value of the dependent variable is explained by the independent variable.

y = 5707.4x + 161189 R2 = 0.1666 $400,000.00 $350,000.00 $300,000.00 $250,000.00 $200,000.00 $150,000.00 $100,000.00 $50,000.00 $0.00 0.00 5.00 10.00 15.00 20.00 25.00 30.00 Unit Cost Of Div 1

Figure 4: R Squared Value In Figure 4, since the plotted points are largely random, the R squared value is near to zero, meaning that only 16% of the variance in the value of the dependent 16

Project Cost

variable is explained by the independent variable. One can think of it as a possibility that any given point will fall on the line described by the regression equation. Coefficient of variance is calculated by dividing the standard deviation by the mean of a data set. It can be used to compare two different data sets. The higher the variance, the higher the variation in the data. For example, if we have a unit cost of division 1, with mean of 21 and standard deviation of 2.5, and the unit cost of division 5 with a mean of 7 and a standard deviation of 0.67, their Coefficients of variance are as follows: Vdiv1 Vdiv5 = 2.5/21 = 0.119 = 0.67/7 = 0.096

So, there is higher variation in the division 1 data. Skewness describes the form of frequency distribution or departure from symmetry for a data set. Positive skew ness indicates a distribution with an asymmetric tail, extending toward more positive values. Negative skewness indicates a distribution with an asymmetric tail, extending toward more negative values. Skewness can be calculated with the following formula: n Skew-ness = xI - xmean 3 (n 1) (n 2) s Outliers An outlier is an extremely high or an extremely low data value when compared with the rest of the values, (Bluman Allan, 1998). The data should be checked for outliers to increase the confidence level on the statistical analysis. Some Statisticians also refer to them as anomalies. Outliers are atypical by definition. They may include infrequent observations or data points, which do not appear to follow the characteristic distribution of the rest of the data. These may appear to reflect genuine properties of the underlying phenomenon (variable), or be due to measurement errors, data input mistakes, or other anomalies, which should not be modeled. (http://www.statsoftinc.com/textbook/gloso.html ) There are several reasons that outliers can be present in a data set. These are: Measurement or observational errors by the person recording the data. Recording or typing error by a person recording the data. Data obtained from a variable, not belonging to the population under consideration. 17

Special cases in the data, whose high or low values are a result of a condition nonexistent in the rest of the population.

There are several ways of identifying outliers in a data population. Some of them are described below: The Grubbs' Method: This method is also called the ESD method (Extreme Studentized Deviate) (http://www.graphpad.com/calculators/GrubbsHowTo.cfm) Calculate the Z-ratio, to find out how far the value to be tested is from the population mean. Mean Value Z= Standard Deviation If the value of Z is large, the value being tested is farther from the rest of the population. There are two cases, which are possible from this point onwards. Case1: The mean and the Standard Deviation are known. Case2: The mean and the standard Deviation are being calculated from the same population including the outlier. Case1: Since 5% of the values in a population are more than 1.96 standard deviations from the mean, it can be concluded that the outlier comes from a different population, if Z is greater than 1.96.

Case2: When calculating the standard deviation and the mean from the given data, the presence of the outlier increases the standard deviation (denominator), and the difference between the mean and the value (numerator). As a result the value of Z does not increase and the outliers cannot be detected. For this purpose Grubb and other statisticians have prepared a table, which identifies the values of Z depending on the population size. If calculated value of Z is more than the value in the table, (Refer Table 1) for a particular population size, then there is 5% chance that the value is not an outlier. The 5% is actually the P-Value, which is (0.05). The P-Value is always inversely proportional to the Z-ratio. Higher Z-ratio, means there are less chances that the value being tested, belongs to population it is being tested against. The P-Value, which gives a percentage for the Z-ratio can be calculated as follows:

18

Using the Table A, we can find the corresponding area under normal distribution for the calculated Z value. For example: assume the Z value to be 2.18, the corresponding area under normal distribution from Table A is 0.4854. Subtract this value from 0.500 to find the area in the right tail. 0.50000 0.4854 = 0.0146, Hence the P Value is 0.0146

N is the population size. N Critical Z N Critical Z N 3 1.15 17 2.62 40 4 1.48 18 2.65 50 5 1.71 19 2.68 60 6 1.89 20 2.71 70 7 2.02 21 2.73 80 8 2.13 22 2.76 90 9 2.21 23 2.78 100 10 2.29 24 2.80 110 11 2.34 25 2.82 120 12 2.41 26 2.84 130 13 2.46 27 2.86 140 14 2.51 28 2.88 15 2.55 29 2.89 16 2.59 30 2.91 Table 1: Corresponding Z values for the population size N. Critical Z 3.04 3.13 3.20 3.26 3.31 3.35 3.38 3.42 3.44 3.47 3.49

This test for outliers can be used online, with the help of a calculator at the following web site: http://www.graphpad.com/calculators/grubbs1.cfm

19

Allen Blumans Method (Bluman Allan, Community College of Allegheny County 1998): Before actually getting into the method to find outliers, it is required to get acquainted with a few terms. Quartile (Qn): Position in fourths of a data value, which is in the distribution. Following is a method to calculate quartile for a population. Consider the data set: 4,5,7,8,9,10,14,15. The example will demonstrate to find the 25th and the 75th percentile of the given data set. 25th Percentile (1st Quartile: - Q1 = (n*p)/100, where n is the population size p is the 25th percentile. Q1 = (8*25)/100 =2 - Since the answer is a whole number 2 nd and 3rd value, which is 5 & 7 in this case, from the data set, in an ascending order will be considered. - Q1= (5 + 7)/2 = 6, so the 25th % or 1st quartile is 6 (Note: Had the value of Q1 been 2.5 (integer) instead of 2 (Whole Number), it would have been rounded off to 3 and the third value from the data set, in an ascending order would have been considered.)

75th Percentile (3rd Quartile3): - Q3 = (n*p)/100, where n is the population size p is the 75th percentile. Q3 = (8*75)/100 =6 - Since the answer is a whole number 6th and 7th value, which is 10 & 14 in this case, from the data set, in an ascending order will be considered. Q3= (10 + 14)/2 = 12, so the 75th % or 3st quartile is 12

The following example will demonstrate, how to calculate the outliers based on the 1st and 3rd Quartiles. Consider the same data set (4,5,7,8,9,10,14,15), - Q1= 6, Q3= 12 Find the Inter-Quartile Range (IQR), which is Q3 - Q1 IQR = Q3 - Q1 = 12 6 = 6 Multiply this value by 1.5 1.5*6 = 9 Subtract this value 9 from Q1 and add it to Q3 Subtract this value 9 from Q1 = 6 - 9 = -3 Add this value 9 from Q3 =12 + 9 = 21

20

Now, check the data set for any values, that fall outside this range 3 to 21. The values that fall outside this range are outliers.

Also, according to the statistician Allen Bluman, when a distribution is normal or bell shaped, data values that are beyond three standard deviations of the mean can be considered as suspected outliers (Bluman Allan , 1998): Wilfrid Dixon (Professor of Bio-mathematics, UCLA) & Frank Massys (Professor of Bio-statistics, UCLA) method of finding outliers. This method is limited to N 20. (N is the population size) Consider four observations 326, 177, 176, 157. The value 326 will be tested to be an outlier in the considered data set. From the Table A on page 72, we can see that r10 is to be used,

X2 - X1 r10 = X4 - X1 177 - 326 r10 = 157 - 326 The value of 95th percentile from the Table A is 0.765. The observation 326 will be rejected, as an outlier, as it is greater than the 95th percentile, which is 0.765. Research & Education Associations method of finding outliers in correlation analysis (Research & Education association , 1996). Standard error: The standard error is a measure of the amount of error, in the prediction of y for an individual x. It can be calculated with the following equation 1 Sy.x = n(n 1) Where n is the population size. x is the independent variable y is the dependent variable 21 ny2 (y)2 -[nxy - (x)(y)]2 nx2 -- (x)2 = 0.882

Those values that fall outside one standard error of the trend line can be considered as outliers. This can be seen in the following graph in Figure 5. The data points outside the two outer parallel lines (which constitute the distance of standard error for the trend line in the center), will be considered outliers.

Chart Title y = 4.2517x + 460.99 R2 = 0.1984 2000 1800 1600 1400 Total Hrs 1200 1000 800 600 400 200 0 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 2 Man Crew Hrs.

Figure 5: Data points outside three standards errors are considered outliers. After the outliers have been identified, care is to be taken while deciding, which data points are to be excluded from the data, and which ones have to be kept. This is a very 22

crucial decision as, care is to be taken, for not excluding the influential data points from the data set. A very clear understanding has to be developed while differentiating between the outliers and the extremities. Method Used To Find Outliers In The Current Data Set At Hand: Step1: Grubbs test for finding outliers was used for the independent variable. The data points identified by this method were tested by their corresponding dependent variables to be extremities or outliers. The detected outliers were then deleted. Step2: Second level of screening was done, by excluding all data points, outside three standard error of the regression line.

23

Parametric Estimation A parametric cost estimate is one that uses Cost Estimating Relationships (CERs), and associated mathematical algorithms (or logic), to establish cost estimates (Scott et. all, 1996). It is important, though, that any CERs used be carefully tested for validity, using standard statistical approaches. The need to reduce the lost bid cost, and reduce the time required for estimating, has led to the parametric cost estimating. Parametric cost estimating is not limited to the construction industry. It is widely used by the manufacturers to estimate the cost of their product. NASA has also used this method of estimation for various purposes. (http://www.jsc.nasa.gov/bu2/) Construction Industry Parametric estimation can be effectively used by the construction industry. Its primary application is to reduce the time involved in bidding a job. To bid a job, those general contractors, who do not self perform any part of the work, call for bids from several subcontractors for each bid item, and then bid the job based on the subcontractors numbers. If the historical data is present and is normalized for multipliers, like inflation factors, and location factors, this historical data can be effectively used to bid a job in considerably less time. One subcontractor can still be called to confirm the accuracy of historical data. Lost bid cost can also be reduced with the help of parametric estimating. These estimates are quite accurate if the historical data is properly captured from its source. When a general contractor bids a job, he spends most of his time in getting the bids from the subcontractor. If this task is reduced to one fourth of its actual duration, the estimator will have more time to cross check other aspects, which are important but are further away on the priority list. In this manner the lost bid cost can be greatly reduced. Parametric estimation can be used at various levels of detail in the construction industry. A construction firm that self performs all, or most of the work, can use this method of estimating in a detailed and accurate manner. This primarily depends on which level they capture their cost data. The levels can be classified as the work break down structure, control accounts, or the sixteen divisions of Construction Specifications Institute (CSI) format. For instance, if the data is captured by the 16 divisions of the CSI format, then it can be recorded at different levels, as explained by the Figure 6, where level five is most detailed and accurate and level 1 is least detailed. Those construction firms that do not self perform the work and hire subcontractors for most of it, can also use parametric estimation techniques if the subcontractors give them a detailed feedback of how many dollars were spent on which item rather than a lump sum amount. This is the primary reason for such general contractors to have a cost accounting history up to Level 1, 2, or a maximum of 3.

24

Div 3 Concrete

Div 4 Metals

Div 5 Woods & Plastics

Structural Metal Framing

Metal Joists

Metal Decking

Cold Formed Metal Framing

Wire Rope Assemblies

Metal Framing System

Steel Core

Fiber Core

Labor

Material

Equipment

Figure 6: Different levels of data capturing in accounting system.

Design Industry

25

Parametric estimation can also be used by the design industry to estimate their design fee when historical data is properly captured. The historical data, in this case would be the time spent by the designer on each phase of the job, (ie. conceptual phase, design phase, etc.). If this data is available, statistical analysis can yield Cost Estimating Relationships (CERs), which in turn would make the design fee estimation process viable. To use parametric estimation techniques in an effective manner, it is necessary to: Have effective cost accounting system Prepare a database from the accounting history. Identify the different multipliers required to normalize the data. Location Factors Inflation Factors Project Type Factors Exclude outliers Identify the important dependent variables Do statistical analysis to find Cost Estimating Relationships. Apply required multipliers. Prepare a template Effective cost accounting system To capture costs in an effective manner, it is important to have a good cost accounting system. Accounting personnel maintain the cost accounting system of an organization, and the pre-construction service personnel use this accounting history to prepare estimates, (in the case of design firms, the designer prepares an estimate of the design fee, based on this accounting history). Due to this reason, there has to be an effective coordination between the accounting system, and the personnel using it to prepare estimates. A good cost accounting system: Is detailed to the required level, and captures the cost data to the required level of detail, of the work breakdown structure. Has clearly classified breakdown structure, Has no repetition of items amongst the cost codes: For instance, if an organization uses CSI format, for their accounting and estimation system. Demolition should not be listed under Division 1 (General Requirements), and Division 2 (Site Construction) as well. This would negatively affect the database, and the estimates prepared from it.

26

Normalize the Data

Database

Accounting History

Exclude Outliers

Exclude Outliers

Statistical Analysis

Statistical Analysis

Closed Job

Detailed Parametric Estimate

Cost Estimating Relationships

Most competent Bidder, Award of the Project

Parametric Estimating Model

Final Estimate Analysis

Adjustment for Project Type Factor

Adjustment for Location Factor

Adjustment for Inflation Factor Figure 7: Parametric Estimating process Estimation and accounting cost codes should match accurately: For instance, if an organization uses CSI format for their accounting and estimation system, cost

27

codes for Brick Masonry should be same in accounting, and estimation system. Incorrect cost codes would obstruct the flow of cost information, and the estimates prepared form such an estimation/ accounting system would be inaccurate. Prepare database from the accounting history. Preparing a database from the accounting history is a very important step in parametric estimating. The fields in a database are primarily items of the work breakdown structure used by the organization. Each field represents the unit cost for an item. The unit can be taken as the square footage of the project, or the time duration of the project. Square footage, as a unit of measure for cost, is more applicable to those parts of the project, where cost is guided by the job size, for example: flooring. Duration of the project, as a unit of measure for cost, is more applicable to those parts of the project, which are guided by the project duration, for example: Division 1 - General Requirements. A wrong data entry or a typing error can make the database less accurate. When doing manual entries to the database in a software like excel, lotus, etc: Crosschecks should be placed at several points, so that if they do not match, the typing error can be identified. If it is not possible to place the crosschecks at frequent data entries, the entries should be crosschecked after entering every 5 10 data points.

It is always advisable to automate the accounting and the estimating system in a manner such that the estimation database is updated with closed jobs at the end of each month, or as frequently as required. If this is not possible then Copy and Paste commands should be used to transfer data from the accounting system, to the estimation system. Care should be taken, not to include in-progress jobs in the database as this would jeopardize the statistical analysis done for estimating. Identify the different multipliers required to normalize the data. Once the database is prepared it needs to be normalized for various factors like location factors, inflation factors, and project type factors. This can be done by using the adjustment factors prepared by the various sources like Engineering News Record, or developing them within the organization. Location Factors: If an organization dose work in various geographical locations and want to use a common database for estimating new jobs, location factors need to be developed, with one state or city as a base with a location factor of 1.0. Cost data from different states, or cities is normalized to this base city. Location factors become important for those organizations, which do not perform a lot of work in one city, to feed a database individually for each city, but perform enough work all over the state, or the country to feed a database.

28

The organization can develop location factors per their requirements. For instance, if they feel that it is not important to be city or state specific, they can divide the geographical region under consideration into four or five parts, like north region, south region, east region and the west region. A combination of the above methods for developing location factors can also be used. Inflation Factors: Those design or construction firms, which do not complete enough jobs to sustain a current database, can use jobs older than one or two years, if they develop inflation factors, thus the old cost data can be normalized to the current year costs. The inflation factors can be developed in-house by an organization, or adopted from other sources like ENR. Of course the in-house developed inflation factors will be more accurate because they are specific to the market in which the project will be built. Inflation factors can be developed on a yearly basis. Project Type Factors: These are usually used when little data is available for estimating a particular project type. In such a case, other projects can be normalized to this project type to populate the database. Project type factors are usually used by organizations that maintain client specific databases. For example, if an estimate has to be done for a particular retail store, and very few jobs for that particular client exist in the accounting history, but there is a good population of other retail stores. A project type factor can be developed. The other retail stores can be normalized to the retail stores under consideration, to have a good population of data for statistical analysis, and in turn an estimate can be done.

Exclude outliers The next step in the process of developing a parametric estimating model is to exclude the outliers from the database. The outliers are of two types: Specific projects, which greatly differ in their costs, when compared to the rest of the projects in the database. For example, small jobs, or specialty alteration jobs. These projects need to be excluded totally from the database Projects with costs of certain items in the work breakdown structure, which is greatly different from the rest of the corresponding items in the database. For these projects only such items should be deleted from the database, and not the whole project.

The specific project outliers, where the whole project has to be deleted, should be filtered before adding closed jobs to the database. The specific item outliers, where an item from the work breakdown structure has to be deleted, should be done after adding the closed jobs to the database, and after doing a preliminary statistical analysis, as these outliers

29

can be detected only upon statistical analysis. Grubbs test can be used for this purpose. For details of Grubbs test, please refer the section of outliers in this research document. Outliers can have a strong effect on the Cost Estimating Relationships, as shown in Figures 8 and 9.
y = 0.0531x + 304.83 R2 = 0.2228 800 700 600 500 400 300 200 100 0 0 1000 2000 3000 4000 5000 6000 7000

Figure 8: A trend line scatter plot, with outliers.

y = 0.106x + 14.113 R2 = 0.9998 800 700 600 500 400 300 200 100 0 0 1000 2000 3000 4000 5000 6000 7000

Figure 9: A trend line scatter plot, without outliers.

30

As can be seen from the figure 8and 9, outliers effect the confidence level of the Cost Estimating Relationship. Confidence level is represented by the R Squared value, which is 0.2228 before excluding outliers, and 0.9998 after excluding outliers. Identify the important dependent variables At this stage it is important to identify the dependent variables for statistical analysis. The dependent variables are the field headings in the database, which yield significant CERs, with the primary dependent variable to be predicted. In the case of design firms, the primary dependent variable is the total design fee, and in case of the construction firms, it is the estimate for a particular job. Identifying the dependent variables, primarily depends on the level of detail, required for the parametric estimate. The higher the level of detail, more accurate is the prediction of he final dollar amount. Once the primary dependent variable is identified, secondary dependant variables are to be identified. Secondary dependant variables are those field headings in the database, which yield a strong CER with the primary dependent variable. The next step is to explore the relationship of all the field headings in the database, with the identified secondary dependent variables. Final Estimate (Primary Dependent Variable) CERs Secondary Dependent Variable 1 Secondary Dependent Variable 2 Secondary Dependent Variable3 CERs Secondary Dependent Variable 4 Secondary Dependent Variable 5

CERs Field Heading 1 of Database Field Heading 2 of Database Field Heading 3 of Database Field Heading 4 of Database Field Heading 5 of Database

Figure 10: Cost Estimating Relationships

31

Do statistical analysis to find Cost Estimating Relationships. After identifying the primary and secondary dependent variables, the next step in the process is to find the cost estimating relationships, of all the field headings with the secondary dependent variables. Those field headings, which yield a strong CER with each of the secondary dependent variables, are used for parametric estimating. Alternative to a Cost Estimating Relationship is a Detailed Parametric Estimate (DEP). This can also be done with the help of statistical analysis, by calculating the measures of central tendency (mean, median, and mode), measures of dispersion (variance, standard deviation, and range), and the measure of confidence (Coefficient of variance, and skewness) for each field on the database. For each item required in the estimate, analysis could be done based on the measures central tendency, dispersion, and confidence level, to finalize unit cost amounts. These unit cost amounts can then be multiplied with the quantities to arrive at a dollar amount for each item. These items can then be added up to arrive at an estimate. Apply required multipliers After finding out the CERs and doing a detailed parametric estimate, the number of different results, or estimates is equal to the number of secondary dependent variables plus, the detailed parametric estimate. All these numbers need to be adjusted to the particular city or state where the project will be built. Incase the database contains data older than two years, it needs to be adjusted to the current year dollars. Also, if required the project type factories are to be used. Prepare a template The final step in the process is to decide, which number is to be used, out of all the CERs and the detailed parametric estimate. The confidence level (R Squared value) associated with each of the CERs, and the detailed parametric estimate, can guide this decision making process. This will be explained with the help of an example (To calculate the design fee for a design firm). Consider that the following CERs, and detailed parametric estimation values were obtained after statistical analysis and application of required multipliers. Amounts yielded by CERs & DPE Confidence Level (R2)

$10000 0.75

$9000 0.79

$10500 0.81

$13000 0.86

$9050 0.90

$12000 0.93

Table 2: Amounts yielded by CERs & DEP Vs their confidence level It can be seen from the

32

It can be seen from Table 2, that the range of the amounts yielded by three highest confidence level CERs/ DEP 0.86, 0.90, and 0.93 are $13000, $9050, and $12000. These include all but one ($9000) readings from the amounts yielded by CERs/ DEP. So, for doing an analysis, we will consider all readings except $9000. These readings can be rated on a five point polar scale, on the basis of their confidence level. Amounts yielded CERs/ DPE Confidence Level (R2) Rating

$10000

$9000

$10500

$13000

$9050

$12000

0.75

0.79

0.81

0.86

0.90

0.93

From Table 3 we can infer that the range of the estimate is $9050 to $13000 and the mean is 10000 + 10500 + 13000 + 9050 + 12000 Mean = 5 = 10910

Standard Deviation = 1582.1 MEDIUM CONFIDENCE $10910

HIGHER CONFIDENCE $9050 - 2 SD - 1 SD

LOWER CONFIDENCE + 1 SD + 2 SD $13000

Incase the estimator is highly confident on the of the number, he can take the lowest number ($9050), in case the level of confidence is medium the mean can be taken ($10919), and if there is not much confident the highest number can be taken ($13000). The main idea is to parametrically estimate the cost using several variables and use professional judgment based on the confidence associated with each number to determine the final value.

33

Methodology The following steps were taken to complete the process of preparing a parametric costestimating database, to estimate design fee, for Multi/ Tech Engineering Services. Data Collection The first step in the process was to collect data from the accounting history of Multi/ Tech Engineering Services. This data was in Lotus 123 format, and was organized project wise. Each project had one or several files for all its data. Care was taken to collect all the required data to save more trips to Salem, Oregon. A contact was also established with personnel from Multi/ Tech for further queries and questions. Preparing The Database The second step was to organize the collected data into various fields to make a database. The database was to be prepared in Microsoft Excel format, so all the 200 Lotus files were converted to Excel format. The fields in the database were the various common aspects amongst all the projects. Each project was individually fed to the database. The database finally comprised of 87 columns and 200 rows, and contained all the information from the 200 Lotus files. Literature Review While preparing the database it was also important to be well acquainted with the concepts of Parametric Estimating and Statistics, so a detailed literature review was done on these topics. The various topics studied under Parametric Estimation were: Preparing a database, Identification of normalization factors, and Dealing with outliers. The topics studied under Statistics were: Measure of central tendency, Measures of dispersion, Scatter plot, Confidence level, and Outliers in regression. Trend Line Graphs The next step was to make trend line graphs, to find out relation between independent and dependent variables. Linear, Logarithmic, and Polynomial Trend lines were plotted. The dependent variables were identified as, The Total Man Hours, The Total Payment, Total Number of Units, and The Total Cost per Unit. The remaining database field headings 34

were considered as independent variables. Total Man Hours, Total Payment, Total Number of Units, and Total Cost per Unit were considered as dependent variables, as their knowledge at the beginning of a project could directly help estimate the total design fee (Total Payment) for the project. Trend line graphs were plotted for each of the dependent variables against each of the independent variables. As a result the graphs were identified in four categories: All Independent variables against Total Man Hours. All Independent variables against Total Payment. All Independent variables against Total Cost per Unit. All Independent variables against Total Number of Units.

The trend line graphs between the dependent variables themselves were also plotted, to estimate the Total Payment. A strong relationship was found between Total Payment and Total Man Hours. Total Cost per Lot, and Total Number of Lots did not yield a significant relationship with either the other dependent variables or the database field headings. It was also important to realize that some relationships, between dependent and independent variables could be specific to different project types. As per recommendations from Multi/ Tech, the database was divided between three project types, and dependent versus independent variables were plotted for each of the project types. The identified project types were: Apartment Projects Subdivision Projects Miscellaneous projects

To summarize, the trend lines graphs were plotted in the following manner: All Projects o All Independent variables against Total Man Hours. o All Independent variables against Total Payment. o All Independent variables against Total Cost per Unit. o All Independent variables against Total Number of Units. Apartment Projects o All Independent variables against Total Man Hours. o All Independent variables against Total Payment. Subdivision Projects o All Independent variables against Total Man Hours. o All Independent variables against Total Payment.

35

Miscellaneous projects o All Independent variables against Total Man Hours. o All Independent variables against Total Payment. Dealing With Outliers

An outlier is an extremely high or an extremely low data value when compared with the rest of the values, (Bluman Allan, 1998). Presence of outliers in a data set makes the statistical analysis inaccurate. It was now important to identify the outliers in the database. The outliers were identified for each of the trend line graphs as, it is important to realize that a project considered to be an outlier in one trend line graph, may not be an outlier in another trend line graph. The outliers were identified primarily with the help of the Grubbs test (Grubbs 1969 and Stefansky 1972). Grubbs test was done on the dependent and the independent variables individually to differentiate between the extremities and the outliers. The next step in identifying outliers was to exclude all those data points on the scatter plot that were out side three standard errors on both sides of the trend lines. It was observed that 70% of the outliers were common for all trend line graphs. Analysis After excluding the outliers, those trend line graphs, which had an R Squared value more than 0.7 were considered strong relationships. Total Payment calculators, were prepared for all project types with the help of these equations. Conclusions and Recommendations Conclusions were drawn and recommendations were made from the analysis.

36

Estimation of Design Fee - Multi Tech Engineering Services Data Collection The first step in the process was to collect data from the accounting history of Multi/ Tech Engineering Services. This data was in Lotus 123 format, and was organized project wise. Each project had one or several files for all its data. Care was taken to collect all the required data to save more trips to Salem, Oregon. A contact was also established with personnel from Multi/ Tech, for further queries and questions. Preparing The Database The second step was to organize the collected data into various fields to make a database. The database was prepared in Microsoft Excel format, so all the 200 Lotus files were converted to Excel format. The fields in the database represent the accounting system organization structure as shown in Figure 11. Each project was individually fed to the database. The database finally comprised of 87 columns and 200 rows, and contained all the information from the 200 Lotus files. The primary dependent variable was identified as the Total Man Hours, and not Total Payment (total design fee), as the Total Payment would be different for the same amount of man-hours for two different years, and an inflation factor would have to be developed, to normalize the historical data. Whereas, total man-hours multiplied by the current rates, could give pre-normalized dollar amounts. The secondary dependent variables were identified as: Total Cost per Unit Total Number of Units Total Payment

Level three independent variables were identified as: Engineer Hrs. Tech III Hrs. Tech II Hrs. Tech I Hrs. AutoCAD Hrs. Plotter Hrs. Clerical Hrs. Mileage Cost Printing Cost Expenses Total Man Hrs. Total Prelm. Hrs. Total As built Hrs. Total Const Hrs. Survey Hours 3 Man Crew Hrs. 2 Man Crew Hrs. Survey Tech Hrs. Reg. Surveyor Hrs. Survey Off. Staff. Hrs.

37

ACCOUNTING ORGANIZATION STRUCTURE OF MULTI/ TECH ENGINEERING SERVICES LEVEL 1 & 2 TOTAL COST PER UNIT TOTAL NUMBER OF UNITS TOTAL MAN HOURS TOTAL MAN HRS. x CURRENT RATES TOTAL PAYMENT

TOTAL DESIGN HOURS TOTAL PRELIM. HOURS

SURVEY

TECH - III TECH - II TECH - I

ENGINEER AUTOCAD PLOTTER MILAGE

3 MAN CREW 2 MAN CREW

SURVEY TECH.

TOTAL ASBUILT HOURS TOTAL CONST. HOURS

REG. SURVEYOR CLERICAL PRINTING

SURVEY OFF. STAFF

EXPENSES

PRELIM. -4

DESIGN

CONST. ASBUILT

Figure 11: Accounting organization structure of Multi/ Tech Engineering Services

38

Level four of independent variables were identified as: Preliminary Hours Design Hours Construction Hours As built Hours Trend Line Graphs The next step was to make trend line graphs to find out relation between independent and dependent variables. Linear, Logarithmic, and Polynomial Trend lines were plotted. The dependent variables were identified as The Total Man Hours, The Total payment, Total Number of Units, and The Total Cost per Unit, and level 3 database field headings were considered as independent variables. Total Man Hours, Total Payment, Total Number of Units, and Total Cost per Unit were considered as dependent variables, as their knowledge at the beginning of a project could directly help, estimate the total design fee (Total Payment) for the project. Trend line graphs were plotted for each of the dependent variables, against each of the independent variables. As a result the graphs were identified in four categories: All Independent variables against Total Man Hours. All Independent variables against Total Payment. All Independent variables against Total Cost per Unit. All Independent variables against Total Number of Unit.

Also the trend line graphs between the dependent variables themselves were plotted, as the final aim was to find out the Total Payment. A strong relationship was found between Total payment and Total Man Hours. Total Cost per Lot, and Total Number of Lots did not yield a significant relationship with either the other dependent variables or the database field headings. It was also important to realize that some relationships, between dependent and independent variables, could be specific to different project types. As per recommendations by Multi/ Tech, the database was divided between three project types, and dependent verses independent variables were plotted, for each of the project types. The identified project types were: Apartment Projects Subdivision Projects Miscellaneous projects

To summarize the trend lines graphs were plotted in the following manner: All Projects o All Independent variables against Total Man Hours. 39

o All Independent variables against Total Payment. Apartment Projects o All Independent variables against Total Man Hours. o All Independent variables against Total Payment. Subdivision Projects o All Independent variables against Total Man Hours. o All Independent variables against Total Payment. Miscellaneous projects o All Independent variables against Total Man Hours. o All Independent variables against Total Payment.

The equations and their R Squared values can be viewed in Appendix II and III Dealing With Outliers It was now important to identify the outliers in the database. The outliers were identified for each of the trend line graphs as, it is important to realize that a project considered to be an outlier in one trend line graph, may not be an outlier in another trend line graph. The outliers were identified primarily with the help of the Grubbs test. Grubbs test was done on the dependent and the independent variables individually, to differentiate between the extremities and the outliers. The next step in identifying outliers was, to exclude all those data points on the scatter plot, that were out side three standard errors, on both sides of the trend lines. It was observed that 70% of the outliers were common for all trend line graphs. Analysis After excluding the outliers, those trend line graphs, which had an R Squared value, more than 0.7 were considered strong Cost estimating Relationships. Total Payment calculators, were prepared for all project types, with the help of these equations. These was a strong relationship observed between Total Payment and Total Man Hours. There were no relationships found between the primary dependent variable, and the other secondary dependent variables, Total Cost per Unit, and Total Number of Units. Other strong relationships observed between primary dependent variable, secondary dependent variables, and Independent variables are as under:

40

All Projects (Y = Total Man Hours) Secondary Relation Independent variable TECH 1 Y = 3.1213X + 104.99 ACAD HRS. Y = 3.2036X + 158.29 PLOTTER Y = 6.9456X + 204.89 3 MAN CREW Y = 14.071X + 380.96 PRINTING COST Y = 0.4636X + 259 TOTAL DESIGN Y = 1.1862X + 132.55 R. Squared Value 0.9097 0.8505 0.7622 0.7186 0.7064 0.8379

SNo. 1 2 3 4 5 6

All Projects (Y = Total Payment, YMEP represents the Normalized Total Payment with mileage, expenses and printing costs and Y represents Normalized Total Payment without mileage, expenses and printing costs). SNo. 1 2 3 4 Secondary Independent variable
TECH I TOTAL SURVEY TOTAL DESIGN TOTAL PAYMENT

Relation
Y = 138.01x + 7367.6 Y = 147.65x + 13268 Y = 56.813x + 6466.6

R. Squared Value
0.7082 0.7167 0.7057 0.9908

YMEP = 1.0707x + 865.4,

All Projects (Y = Total Payment, YMEP represents the Normalized Total Payment with mileage, expenses and printing costs and Y represents Normalized Total Payment without mileage, expenses and printing costs). SNo. 1 2 Secondary Dependent variable Total Man Hours Total Man Hours Relation Y = 47.841x - 457.26 YMEP= 51.856x - 3.926 R. Squared Value 0.9477 0.9409

Ap

41

Apartment Projects (Y = Total Man Hours) Secondary SNo. Relation Independent variable
1 2 3 4 5 6 7 ENGINEER TECH 1 ACAD HRS. PLOTTER 3 MAN CREW PRINTING COST TOTAL DESIGN Y = 8.6459X + 186.59 Y = 2.955X + 85.256 Y = 3.1146X + 93.24 Y = 7.4265X + 72.492 Y = 16.211X + 278.09 Y = .3988X + 179.09 Y = 1.214X + 71.316

R. Squared Value
0.7234 0.9223 0.9235 0.8014 0.8173 0.8304 0.8818

Apartment Projects (YMEP = Total Payment including Mileage, Electrical, and Printing costs) Secondary R. Squared SNo. Relation Independent variable Value YMEP = 127.65x + 7048.8 1 TECH I 0.8063 Y = 134.93x + 7199.1 2 ACAD HRS 0.7788 MEP YMEP = 294.63x + 5856 3 PLOTTER HRS 0.7477 YMEP = 938.27x + 13997 4 3 MAN CREW 0.7129 YMEP = 54.741x + 6060.1 5 TOTAL DESIGN HRS 0.8315 6 TECH I DESIGN HRS. YMEP = 139.73x + 9714.8 0.7636 Apartment Projects (Y = Total Payment, YMEP represents the Normalized Total Payment with mileage, expenses and printing costs and Y represents Normalized Total Payment without mileage, expenses and printing costs). SNo.
1 2

Level 3 Independent Relation variable Total Man Hours Y = 42.21x + 612.88 Total Man Hours YMEP.= 47.417x + 822.19

R. Squared Value
0.9765 0.9546

Miscellaneous Projects (Y = Total Man Hours) Level 3 Independent variable

ENGINEER TECH II TECH 1 ACAD HRS. PLOTTER PRINTING COST

SNo.
1 2 3 4 5 6

Relation
Y = 8.5322X + 172.43 Y = 4.7654X + 216.29 Y = 2.9786X + 80.893 Y = 2.9458X + 122.59 Y = 8.7197X + 87.102 Y = 0.5787X + 164.54

R. Squared Value
0.8144 0.7144 0.9516 0.876 0.901 0.8579

42

7 8

TOTAL DESIGN TOTAL CONST.

Y = 1.1132X + 105.81 Y = 1.7212X + 213.93

0.801 0.7067

Miscellaneous Projects (Y = Total Payment) Level 3 Independent variable

ENGINEER TECH I ACAD HRS PRINTING EXPENSES TOTAL DESIGN TOTAL CONST.

SNo.
1 2 3 4 5 6 7

Relation YMEP = 338.01x + 11936 YMEP = 123.85x + 5394 YMEP = 123.65x + 6830.4 YMEP = 26.602x + 8434.8 YMEP = 27.906x + 13006 YMEP = 43.32x + 8802.1 YMEP = 53.514x + 10854

R. Squared Value
0.7975 0.8511 0.8121 0.8477 0.7226 0.754

0.7602

Miscellaneous Projects (Y = Total Payment, YMEP represents the Normalized Total Payment with mileage, expenses and printing costs and Y represents Normalized Total Payment without mileage, expenses and printing costs). SNo.
1 2

Level 3 Independent Relation variable Total Man Hours Y = 43.318x + 654.57 Total Man Hours YMEP.= 47.132x + 633.74

R. Squared Value
0.9552 0.9416

Subdivision Projects (Y = Total Man Hours) Level 3 Independent variable

TECH I AUTOCAD PLOTTER HRS PRINTING COST TOTAL DESIGN HRS

SNo.
1 2 3 4 5

Relation
Y = 3.216X + 125.26 Y = 3.337X + 199.42 Y = 6.2839X + 300.98 Y = 0.5033X + 278.08 Y = 1.1268X + 199.49

R. Squared Value
0.9068 0.8528 0.7781 0.7002 0.8152

Subdivision Projects (Y = Total Payment, YMEP represents the Normalized Total Payment with mileage, expenses and printing costs and Y represents Normalized Total Payment without mileage, expenses and printing costs). SNo.
1 2

Level 3 Independent variable Total Man Hours Total Man Hours

Relation Y = 51.652x - 1701.7 YMEP= 53.848x - 727.29 43

R. Squared Value
0.9636 0.9537

Subdivision Projects (Y = Total Man Hours) Level 3 Independent variable

TECH I ACAD TOTAL SURVEY TOTAL DESIGN TOTAL PAYMENT

SNo.
1 2 3 4 5

Relation Y = 143.11x + 10140 Y = 164.85x + 11025 Y= 148.65x + 12122 Y = 53.419x + 11610 YMEP = 1.074x + 589.74

R. Squared Value
0.7404 0.7166 0.8222 0.714 0.9945

Division of Man Hours into Different Categories. SNo. All Projects

8.05% 2.98% 11.68% 26.91% 22.90% 9.79% 2.40% 15.29% 100.00%

Apt. Projects Misc. Projects Sub. Projects

7.41% 3.11% 6.61% 29.88% 27.37% 11.78% 1.86% 11.98% 100.00% 7.00% 4.68% 11.27% 26.89% 23.35% 9.27% 3.13% 14.42% 100.00% 7.79% 1.86% 12.57% 24.54% 20.85% 8.56% 1.81% 22.03% 100.00%

Engineer Tech III Tech II Tech I AutoCAD Plotter Clerical Survey Total

Division of Survey Hours into Different Categories. SNo. All Projects Apt. Projects Misc. Projects Sub. Projects
12.73% 46.93% 26.35% 11.47% 2.52% 100.00% 8.37% 53.75% 21.14% 14.99% 1.75% 100.00% 17.67% 43.85% 17.96% 18.22% 2.30% 100.00% 12.78% 45.33% 31.05% 7.95% 2.88% 100.00%

3 Man Crew 2 Man Crew Reg. Surveyor Survey Tech Survey off. Staff Total

Separate calculation templates were prepared for, all projects, apartment projects, miscellaneous projects, and subdivision projects. Those relationships between dependent variables and independent variables, which had a higher R Squared value, were considered stronger than those with lower R Squared values.

44

45

46

47

48

Testing the Parametric Cost Estimating Model An apartment project was chosen to test the parametric cost-estimating model. The primary independent variables have been identified as Engineer Design Hours, Tech I Design Hours, AutoCAD Design Hours, and Plotter Design Hours. These will have to be estimated, as the first step in parametrically estimating the design fee. The secondary independent variables can be calculated with the knowledge of these four primary independent variables, by using the equations in Table 3.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Y Predicting Total AutoCAD Hrs. Predicting Total AutoCAD Hrs. Predicting Total AutoCAD Hrs. Predicting Total Design Hrs Predicting Total Design Hrs Predicting Total Design Hrs Predicting Total Design Hrs Predicting Total Mileage Cost Predicting Total Mileage Cost Predicting Total Printing Cost Predicting Total Printing Cost Predicting Total Survey Hrs Predicting Total Survey Hrs Predicting Total Engineer Hrs. Predicting Total Tech I Hrs. Predicting Total Tech I Hrs. Predicting Total Tech I Hrs. Predicting Total Tech II Hrs. Predicting Total Tech III Hrs. Predicting Total Tech III Hrs. Predicting Total Plotter Hrs. Predicting Total Plotter Hrs.

X AutoCAD Design Hrs. Tech I Design Hrs Total Tech I Hrs. Tech I Design Hrs Tech I Hrs AutoCAD Design Hrs. Total AutoCAD Hrs. Tech III Design Hrs. Total Tech III Hrs. Tech I Design Hrs. Tech I Hrs 2 man Crew 3 Man Crew Design Engineer Hrs. AutoCAD Design Hrs. Tech I Design Total AutoCAD Hrs Tech II Const. Hrs. Tech III As built Hrs Tech III Const. Hrs. Plotter Design Hrs Total Tech I Hrs.

RELATION y = 1.073x + 14.678 y = 0.9777x + 28.896 y = 0.8864x + 12.621 y = 2.3971x + 82.294 y = 1.9865x + 73.758 y = 2.4642x + 70.233 y = 2.1238x + 64.401 y = 30.274x + 648.49 y = 16.898x + 526.84 y = 5.8763x + 183.26 y = 5.4171x + 122.72 y = 1.5591x + 11.369 y = 3.4415x + 29.317 y = 1.3131x + 6.7633 y = 1.0649x + 22.248 y = 1.0961x + 19.297 y = 1.0115x + 4.5963 y = 0.989x + 20.951 y = 2.1249x + 15.251 y = 1.784x + 4.1796 y = 1.1327x + 4.8069 y = 0.2778x + 20.937

R2 0.9273 0.8334 0.8966 0.9348 0.8403 0.9125 0.8417 0.8892 0.8733 0.7556 0.792 0.7562 0.7071 0.821 0.8003 0.9179 0.8966 0.7643 0.9763 0.9781 0.8616 0.7097

Table 3 After calculating the secondary independent variables (Engineer Hours, Tech I Hours, AutoCAD Hours, Plotter Hours, Printing Cost, and Total Design Hours), as shown in Table 4 the next step is to find out their relationship with the secondary dependent variable, the Total Man Hours.

49

ENGINEER 175

TECH 1 562

ACAD HRS. 521

PLOTTER 187

PRINTING COST 3640.5

TOTAL DESIGN 1503

Table 4 Based on the equations in Table 5, Total Man Hours can be calculated: Independent Variable
ENGINEER TECH 1 ACAD HRS. PLOTTER PRINTING COST TOTAL DESIGN

Relationship with Total Man Hours

Y = 8.6459X + 186.59 Y = 2.955X + 85.256 Y = 3.1146X + 93.24 Y = 7.4265X + 72.492 Y = .3988X + 179.09 Y = 1.214X + 71.316

R2
0.7234 0.9223 0.9235 0.8014 0.8304 0.8818

Result
1699.62 1745.97 1715.95 1461.25 1630.92 1895.96

Table 5 Division of Total Man Hours into various categories was calculated based on results of Table 5 and the apartment project breakup percentages. Apt. Projects Engineer %
7.41% 3.11% 6.61% 29.88% 27.37% 11.78% 1.86% 11.98% 100.00% 0.7234 125.94 52.86 112.35 507.85 465.19 200.22 31.61 203.61 1699.62

SNo. R2
Engineer Tech III Tech II Tech I AutoCAD Plotter Clerical Survey Total

Tech 1 Acad Hrs. Plotter

0.9223 129.38 54.30 115.41 521.69 477.87 205.67 32.47 209.17 1745.97 0.9235 127.15 53.37 113.42 512.72 469.65 202.14 31.92 205.57 1715.95 0.8014 108.28 45.44 96.59 436.62 399.94 172.13 27.18 175.06 1461.25

Printing Cost
0.8304 120.85 50.72 107.80 487.32 446.38 192.12 30.34 195.38 1630.92

Total Design
0.8818 140.49 58.96 125.32 566.51 518.92 223.34 35.26 227.14 1895.96

Table 6 Survey hrs from Table 6 were divided into 3 man crew, 2 man crew, registered surveyor, survey tech, & survey office staff based on their percentages arrived at by statistical analysis.

50

Survey Staff 3 Man Crew 2 Man Crew Registered Surveyor Survey Tech Survey Off Staff Total

% of Total Survey Hrs. R2 8.37% 53.75% 21.14% 14.99% 1.75% 100.00%

Engineer 0.7234 17.05 109.43 43.05 30.52 3.56 203.61

Tech 1 0.9223 17.51 112.42 44.22 31.36 3.66 209.17

Acad Hrs.
0.9235 17.21 110.48 43.46 30.82 3.60 205.57

Plotter 0.8014 14.66 94.08 37.01 26.24 3.06 175.06

Printing Cost 0.8304 16.36 105.01 41.31 29.29 3.42 195.38

Total Design 0.8818 19.02 122.07 48.02 34.05 3.97 227.14

Table 7 Hourly Rate Table (From/ Multi Tech Engineering Services)

Engineer $95.00 Clerical $35.00 TECH III $60.00 3 Man Crew $105.00 Tech II $52.50 2 Man Crew $77.50 Tech I $45.00 Registered Surveyor $65.00 AutoCAD $27.50 Survey Tech. $55.00 Plotter $11.50 Survey Off Staff $42.00

Table 8

Survey Staff 3 Man Crew 2 Man Crew Registered Surveyor Survey Tech Survey Off Staff Total

% of Total Survey Hrs. R2 $105.00 $77.50 $65.00 $55.00 $42.00 $344.50

Engineer 0.7234 1789.93 8481.07 2798.19 1678.77 149.63 14897.58

Tech 1 0.9223 1838.74 8712.32 2874.49 1724.54 153.71 15303.79

Acad Hrs.
0.9235 1807.12 8562.52 2825.06 1694.89 151.07 15040.67

Plotter 0.8014 1538.89 7291.58 2405.74 1443.32 128.64 12808.17

Printing Cost 0.8304 1717.58 8138.25 2685.08 1610.91 143.58 14295.40

Total Design 0.8818 1996.70 9460.78 3121.43 1872.69 166.92 16618.51

Table 9

Finally the Total Payment was calculated based on Tables 7, 8 and 9. SNo. R2 Apt. Printing Engineer Tech 1 Acad Hrs. Plotter Projects % Cost
0.7234 0.9223 0.9235 0.8014 0.8304

Total Design
0.8818

51

Engineer Tech III Tech II Tech I AutoCAD Plotter Clerical Survey Total

$95.00 $60.00 $52.50 $45.00 $27.50 $11.50 $35.00

11964.49 3171.50 5898.11 22853.12 12792.63 2302.48 1106.45 14897.58 74986.38

12290.73 3257.97 6058.94 23476.26 13141.45 2365.26 1136.62 15303.79 77031.02

12079.41 3201.96 5954.76 23072.62 12915.50 2324.59 1117.08 15040.67 75706.59

10286.45 2726.69 5070.89 19647.93 10998.44 1979.55 951.27 12808.17 64469.41

11480.87 3043.30 5659.70 21929.37 12275.54 2209.41 1061.73 14295.40 71955.32

13346.60 3537.86 6579.45 25493.05 14270.40 2568.45 1234.27 16618.51 83648.59

Table 10 Other important relationships found with Total Payment (Y = Total Payment, YMEP represents the normalized Total Payment with mileage, expenses and printing costs and Y represents normalized Total Payment without mileage, expenses and printing costs).

Independent Variable
TOTAL MAN HRS. TOTAL PAYMENT WITHOUT MEP TOTAL MAN HRS. TECH I ACAD HRS PLOTTER HRS 3 MAN CREW TOTAL DESIGN HRS TECH I DESIGN HRS.

Relationship with Total Payment

Y = 42.21x + 612.88 YMEP = 1.1287x + 2.0285 YMEP = 47.417x + 822.19 YMEP = 127.65x + 7048.8 YMEP = 134.93x + 7199.1 YMEP = 294.63x + 5856 YMEP = 938.27x + 13997 YMEP = 54.741x + 6060.1 YMEP = 139.73x + 9714.8

R2
0.9765 0.9868 0.9546 0.8063 0.7788 0.7477 0.7129 0.8315 0.7636

Table 11

Total Payment calculated based on Table 6, 10, & 11.

CALCULATING TOTAL PAYMENT TOTAL MAN HRS. , YMEP TOTAL PAYMENT FROM TABLE 11 0.9748
$78,788.10

R2 Engineer Tech 1
0.7234 0.9223

0.9546 $81,413.07 $83,610.85

R2
0.8063

52

AutoCAD Hrs. Plotter Printing Cost Total Design

Tech I Design Hrs

0.9235 0.8014 0.8304 0.8818

$82,187.39 $70,110.28 $78,155.52 $90,722.93

$77,497.63 $60,951.81 $88,335.82 $81046.97

0.7788 0.7477 0.8315 0.7636

Table 12

TOTAL PAYMENT PRIDECTED BY OTHER VARIABLES

$100,000.00 $90,000.00 $80,000.00 TOTAL PAYMENT $70,000.00 $60,000.00 $50,000.00 $40,000.00 $30,000.00 $20,000.00 $10,000.00 $0.00
72 % 76 % 77 % 80 % 83 % 84 % 85 % 86 % 86 % 88 % 88 % 89 % 92 % 92 % 92 % 94 % 94 %

R SQUARED VALUES

Figure 12: Bar chart showing the Predicted Total Payment for varying R2 Values

S. NO.
1 2 3 4 5

SECONDARY INDEPENDENT SECONDARY/ PRIMARY MEAN R2 VARIABLE DEPENDENT VARIABLE RESULT VALU 2 2 1R 2R E
Plotter Printing Cost Engineer Acad Hrs. Tech 1 80.14% 83.04% 72.34% 92.35% 92.23% YMEP YMEP YMEP YMEP YMEP 95.46% 95.46% 95.46% 95.46% 95.46% 87.80% 89.25% 83.90% 93.91% 93.85% $70,110.28 $78,155.52 $81,413.07 $82,187.39 $83,610.85

53

6 7 8 9

Total Design TECH I DESIGN HRS PLOTTER ACAD HRS.

88.18% 80.14% 92.35% 92.23% 88.18% 80.14% 83.04% 72.34% 92.35% 92.23%

YMEP Y Y Y Y Y

95.46% 76.36% 74.77% 77.88% 80.63% 83.15%

91.82% 76.36% 77.46% 85.12% 86.43% 85.67%

$90,722.93 $81,046.97 $60,951.81 $77,497.63 $78,788.10 $88,335.82 $64,469.41 $71,955.32 $74,986.38 $75,706.59 $77,031.02

10 TECH 1 11 TOTAL DESIGN 12 PLOTTER 13 PRINTING COST 14 ENGINEER 15 ACAD HRS. 16 TECH 1

RATE 92.35% RATE 92.23%

RATE 88.18%

$83,648.59

17 TOTAL DESIGN

88.18% AVERAGE MEDIAN MODE STANDARD DEVIATION VARIANCE RANGE COEFFICIENT OF VARIANCE SKEWNESS HIGHEST VALUE LOWEST VALUE 85.87% 86.43% #N/A 77683.39 78155.52 #N/A

MEASURE OF CENTRAL TENDENCY

MEASURE OF DISPERSION

0.06 7736.02 0.00 59845952.16 21.57% 29771.12 0.07 -0.63 0.10 -0.55

MEASURE OF CONFIDENCE LEVEL

LARGEST & SMALLEST VALUE

93.91% $90,722.93 72.34% $60,951.81

Table 13 Analysis Since there is a considerable amount of variance, a high standard deviation, and a large Range in the data, the average of all the available results should not be taken. PREDECTING TOTAL PAYMENT
VALUES ASSOCIATED WITH R2

AVERAGE R2 MEDIAN R2
86.43% 85.67%

AVERAGE RESULT

MEDIAN RESULT

54

HIGH------------------ AVERAGE----------------------LOW

+ ,+ ,+ ,+ ,+
,

* %* %* %* %*
%

RATE 80.14% RATE 83.04% RATE 72.34%

$78,788.10 AVERAGE (FINAL NUMBER)

$88,335.82 $77,683.39 $80,740.7075

$78,155.52

Table 14 This process can be summarized in the following steps: Step 1: Identify the required secondary independent variables from Section 2 of the template. Step 2: Identify the required primary independent variables in Section 1 of the template, based on Step 1 and estimate them. Step 3: Find the values of secondary independent variables (Engineer, Tech I, etc.) with the help of the equations in Section 1 of the template. If there is more than one value for any secondary independent variable, take the mean or choose one of the values based on their R2 values. Step 4: Based on the equations in Section 2 of the template, calculate the secondary dependent variable. This will yield eight values for Total Man Hours (secondary dependent variable). Calculate the confidence level for each of these eight values by taking a mean of the confidence levels of the two equations previously used. Step 5: Divide all these eight values individually into Engineer Hrs, Tech III Hrs, etc., based on the percentages given in Section 3 of the template. Their values can be compared with those calculated in Step 3. If the values calculated in Step 3 have a higher R2 values, they can be adopted as well. This will primarily depend on the estimators analysis at this stage. Step 6: Multiply the values obtained from Step 5 with the current rates to get 8 or 9 different values for the Total Design Fee. Their confidence level will be the same as that in Step 4. Step 7: Also calculate the Total Payment based on the equations in Section 4 of the template and the values obtained from Step 4. This will yield 8 values for Total Payment. Their confidence level will be the mean of the R2 values of the equations used from Section 4, and those calculated in Step 4. At this stage we have sixteen different results for Total Payment with their individual confidence levels. Step 8: Use the equations in Section 5 and the values calculated with the equations in Section 1 of the template to arrive at 6 more values of Total Payment. Their level of confidence will be the mean of R2 values the equations used from Section 5 and of arrived at in Step 3. At this stage we have 22 values for Total Payment and their associated confidence levels.

55

PRIMARY INDEPENDENT VARIABLE

SECONDARY INDEPENDENT VARIABLE

SECONDARY DEPENDENT VARIABLE Path A Path B Path C R2 = Constant PRIMARY DEPENDENT VARIABLE

Figure 13: Confidence on the primary dependent variable: Longer the relationship, lower the confidence level. Step 9: To decide which value to count on most or to take a mean of all the values, is a very curtail decision. This depends on the Measures of Central Tendency, Measures of Dispersion, Measure of Confidence Levels and the R2 values, as shown in Table 13. If there is a strong relation between primary independent variable and the primary dependent variable, as shown by Path A in Figure 13, the number obtained from such a relation should be given maximum wightage. The number obtained from Path C (if the confidence level is same as that for Path A) should be given minimum weightage. The thumb rule is (with R2 constant): Longer the relationship lower is the level of confidence. Conclusions and Recommendations Tech I and AutoCAD hours have a strong CER with many secondary independent variables.

56

Doing a detailed estimate of Tech I and AutoCAD hours (Design, Construction, and As built hours), can lead to a accurate estimation of the Total Payment. There was no CER found between Total Man Hours and Mileage cost or Printing cost. Total Man Hours have slight difference of percentage division amongst Engineer Hours, Tech III Hours, etc., when compared across Apartment, Miscellaneous, & Subdivision Projects. Total Survey Hours have slight difference of percentage division amongst 3 Man Crew Hours, 2 Man Crew Hours, etc., when compared across Apartment, Miscellaneous, & Subdivision Projects. Total Payment excluding mileage, printing, and expenses costs have stronger CER (higher R2 values) with the secondary dependent variables like Engineer hrs, Tech I hrs., etc., as compared to Total Payment including mileage, printing, and expenses costs. Total Payment excluding mileage, printing, and expenses costs has a strong CER (high R2 values) with Total Payment including mileage, printing, and expenses costs. Tech I, AutoCAD, and Total Design Hours have a strong CER with Total Payment across all project types. When calculating Total Payment (including mileage, printing, and expenses cost) with the help of an equation with R2 value in 0.9 to 1.0 range, 90% of the cases yield a value within 5% of the actual number. When calculating Total Payment (including mileage, printing, and expenses cost) with the help of an equation with R2 value in 0.8 to 0.9 range, 90% of the cases yield a value within 12 - 15% of the actual number. Total Payment had no CER with the Total Number of Units or Cost Per Unit in any project type. Limitations of the Database

The primary independent variables are those variables that are known at the beginning of a job, these are the variables that can yield an accurate CER with the Primary Dependent Variable. In the case under consideration, such variables were unknown, so other variables like Tech I Design Hours, AutoCAD Design Hours, etc. were considered as primary independent variables. These will have to be estimated before using the templates, which have been developed for parametric estimating. A good example of a primary independent variable is square footage of the lot. The database contains actual values of man-hours or cost, whereas it would have been much better if these were unit costs or unit hours, as with the knowledge of unit values a detailed parametric estimate could have been done, which would compliment the CERs and intern increase the level of confidence on the final number. 57

Areas for Future Research A parametric estimating template based on the cost estimating relationships need to be tested with several projects which are not included in the database. This makes the estimator confident of the CERs or identifies those CERs, which, in spite of having a high R2 value do not yield accurate numbers. So it is important to further test these templates with new projects. It is possible to write a program in languages like Pearl or C++, which is linked to the database and keeps on updating the CERs and the parametric estimating template as new projects are added to the database. It is also possible to prepare a calculator, which does calculations based on the CERs and yields a final estimate upon entering the primary independent variables. The current database can be greatly enhanced if the variables like lot size and project duration are known. New CERs can be found out to enhance the parametric estimating templates.

58

Bibliography Barnett and Lweis. (1994). Outliers in Statistical Data: J. Wiley, New York. Blischke, Wallace R. and Murthy Prabhakar D.N. (2000). Reliability Modeling, Prediction, and Optimization: John Wiley, New York. Charette, Robert P. and Harold E. Marshall. (1999). UNIFORMAT II Elemental Classification for Building Specifications, Cost Estimating, and Cost Analysis: U.S. Dept. of Commerce, Technology Administration, National Institute of Standards and Technology. Chatterjee, Sampreet. (1988). Sensitivity Analysis in Linear Regression: J. Wiley, New York. Cunningham, Marthe. (1977). Integrated Cost Solutions. [WWW document]. URL http://www.pricesystems.com/services/main.html Dixon, Wilfrid, and Massey Frank (1969). Introduction to Statistical Analysis: McGraw-Hill Book Company, New York. Gallagher, Paul F. (1982). Parametric Estimating for Executives and Estimators: Van Nostrand Reinhold, New York. Genichi, Taguchi. (1987). System of Experimental Design Engineering methods to optimize quality and minimize costs: UNIPUB/Kraus International Publications Dearborn, Mich. Kent, Davis and W.B. Ledbetter. (1987). Measuring Design and Construction Quality Costs: Construction Industry Institute, the University of Texas at Austin Marple, S.Lawrence Jr. (1987). Digital Signal Analysis Techniques. [WWW document]. URLhttp://www.cei.se/011.htm. McCulloch, and Charles E. (2001). Generalized, Linear, and Mixed Models: J. Wiley, New York. McLachlan, G.J. and Peel D. (2000). Finite Mixture Models: John Wiley & Sons, New York. Mechefske, C.K. (1995). Auto Regressive Moving Average Estimation. [WWW document]. URL http://www.wavelet.org/wavelet/tutorial/pse.htm.

59

Research and Education Association. (1996). The Statistics Problem Solver: Research and Education Association, Piscataway, New Jersey. Rodriguez, German. (1986). Statistical Demography. [WWW document]. URL http://opr.princeton.edu/courses/pop501/. Ryan, Thomas P. (2000). Statistical Methods for Quality Improvement: Wiley, New York. Schimek, Michael G. (2000). Smoothing and Regression, Approaches, Aomputation, and Application /: J. Wiley, New York. Spector, Eleanor A. (1996). Parametric Techniques Calibration and Validation. [WWW document]. URL http://www.dtic.mil/ndia/tacom/roche.pdf. Spector, Eleanor A. (1998). Parametric Cost Estimating Handbook. [WWW document]. URL http://www.ispa-cost.org/PEIWeb/newbook.htm. Stasiowski, Frank A., Burstein David. (1994). Total Quality Project Management For The Design Firm: J. Wiley, New York. Stat Point Corporate Center. Online Calculator for Probability Distribution. [WWW document]. URL: http://www.statlets.com/free/pdist.htm Therrien. (1989). Decision Estimation and Classification. [WWW document]. URL http://www.cs.ucsb.edu/~yfwang/courses/cs595/99summer/. Walker, Frank R. The Building Estimator's Reference Book; A Practical and Thoroughly Reliable Reference Book for Contractors and Estimators Engaged in Estimating the Cost of and Constructing all Classes of Modern Buildings: Frank R. Walker company, Chicago. Wasserman. (1989). Logistic Regression. [WWW document]. URL http://www.statsoftinc.com/textbook/gloso.html. Williams, John. (1983). Cost Estimating and Value Engineering. [WWW document]. URL http://www.uscost.com/services.htm.

60

APPENDIX 1 Definitions of Terms (Parametric Cost Estimating Handbook by: Robert Scott, Robert Spiker, and Michael Thibault, 1996) Algorithmic Models - (also known as Parametric models) produce a cost estimate using one or more mathematical algorithms using a number of variables considered to be the major cost drivers. These models estimate effort or cost based primarily on the Hardware/Software size, and other productivity factors known as cost driver attributes. Analogy - A method of estimating developed on the basis of similarities between two or more programs, systems, items, etc. Analogy Models - use a method of estimating that involves comparing a proposed project with one or more similar and completed projects where costs and schedules are known. Then, extrapolating from the actual costs of completed projects, the model(s) estimates the cost of a proposed project. Analometric - A recent term meaning a method of combining the analogy and parametric estimating methods to form a cost estimate when only 2 relevant data points are available. It is usually combined with the "D" Factor to adjust (up or down) complexity from the 2 point CER regression line. Analysis - decision making context involving time horizons extending into the future. A concrete set of specifications are not usually available. There could be many major uncertainties, and a wide range of alternatives, each having several configurations. Analysis is usually concerned with new equipment proposals and new methods of operation of systems never produced before. Analysis Objectives are to find significant differences in resource requirements among alternatives; and, how will resource requirements for any alternative change as key configuration characteristics vary over their relevant ranges. It is often a "sensitivity" type of investigation. Analysis (NES Dictionary) - a systematic approach to problem solving. Complex problems are simplified by separating them into more understandable elements. Anomalies - variances in cost related data caused by an unusual event(s) not expected to recur in the future. As Spent Dollars - the cost, in real year dollars, of a project recorded as the dollars were spent without normalization for inflation. Bottom-Up Models - use a method of estimation that estimates each component of the project separately, and the results are combined ("Rolled Up") to produce an estimate of the entire project. Calibration - in terms of a cost model, a technique used to allow application of a general model to a specific set of data. This is accomplished by calculating adjustment factor(s) to compensate for differences between the referenced historical costs and the costs predicted by the cost model using default values. 61

Constant Dollars - Computed values which remove the effect of price changes over time (inflation). An estimate is said to be in constant dollars if costs for all work are adjusted to reflect the level of prices of a base year. Cost Analysis (NES Dictionary) - the accumulation and analysis of actual costs, statistical data, and other information on current and completed contracts or groups of contracts (programs). Cost analysis also includes the extrapolation of these cost data to completion, comparisons and analyses of these data, and cost extrapolations on a current contract with the cost data in the contract value for reports to customers, program and functional managers, and price estimators. In the procurement organizations of the Department of Defense, cost analysis is the review and evaluation of a contractor's cost or pricing data and of the judgmental factors applied projecting from the data to the estimated costs, in order to form an opinion on the degree to which the contractor's proposed costs represent what performance of the contract should cost, assuming reasonable economy and efficiency. Cost Analysis (Large) - the primary purpose of cost analysis is comparison - to provide estimates of the comparative or relative costs of competing systems, not to forecast precisely accurate costs suitable for budget administration. In this context consistency of method is just as important, perhaps more so, as accuracy in some absolute sense. In comparing the costs of military systems, we prefer to speak of "cost analysis" rather than "cost estimation," because the identification of the appropriate elements of cost -- the analytical breakdown of many complex interrelated activities and equipment -- is so important a part of the method. Weapon system cost analysis is much more than an estimate of the cost of the weapon itself. Weapon procurement costs may be relatively small compared to other necessary costs, such as base facilities, training of personnel, and operating expenses; and these other costs may vary greatly from system to system. Cost Benefit Analysis (NES Dictionary) - a technique for assessing the range of costs and benefits associated with a given option, usually to determine feasibility. Costs are generally in monetary terms. Cost Drivers - The controllable system design or planning characteristics that have a predominant effect on the system's costs. Those few items, using Pareto's law, that have the most significant cost impact. Cost Effectiveness (NES Dictionary) - the measure of the benefits to be derived from a system with cost as a primary or one of the primary measures. Cost Effectiveness Analysis (NES Dictionary) - a method for examining alternative means of accomplishing a desired military objective/mission for the purpose of selecting weapons and forces which will provide the greatest military effectiveness for the cost. Cost Estimating (Fisher) - "making an estimate" of the cost of something implies taking a rather detailed set of specifications and "pricing them out." Cost Estimating (NES Dictionary) - cost and price estimating is defined as the art of predetermining the lowest realistic cost and price of an item or activity which assure a normal profit. 62

Cost Estimating Relationship - An algorithm relating the cost of an element to physical or functional characteristics of that cost element or a separate cost element; or relating the cost of one cost element to the cost of another element. Cost Estimating Relationships: CER - a mathematical expression which describes, for predicative purposes, the cost of an item or activity as a function of one or more independent variables. Cost Factor - a brief arithmetic expression wherein cost is determined by the application of a factor as a proportion. Cost Model - an estimating tool consisting of one or more cost estimating relationships, estimating methodologies, or estimating techniques used to predict the cost of a system or one of its lower level elements. Cost/Schedule Control System Criteria: C/SCSC - a set of criteria specified by the Federal Government for reporting project schedule and financial information. Current Dollars - Level of costs in the year actual cost will be incurred. When prior costs are stated in current dollars, the figures given are the actual amounts paid. When future costs are stated in current dollars, the figures given are the actual amounts expected to be paid including any amount due to future price changes. Deflators - The de-escalation factors used to adjust current cost/price to an earlier base year for comparison purposes. A deflator is the inverse of an escalator. Delivered Source Instructions: DSIs - the number of source lines of code developed by the project. The number of DSIs is the primary input to many software cost estimating tools. The term DELIVERED is generally meant to exclude non-delivered support software such as test drivers. The term SOURCE INSTRUCTIONS includes all program instructions created by project personnel and proceed into machine code by some combination of preprocessors, compilers, and assemblers. It excludes comments and unmodified utility software. It includes job control language, format statements, and data declarations. Detail Estimating: Grass Roots, Bottom-Up - The logical buildup of estimated hours and material by use of blue-prints, production planning tickets, or other data whereby each operation is assigned a time value. Design to Cost: DTC - An acquisition management technique to achieve defense system designs that meet stated cost requirements. Cost is addressed on a continuing basis as part of a system's development and production process. The technique embodies early establishment of realistic but rigorous cost objectives, goals, and olds and thresholds and a determined effort to achieve them. Cost, as a key design parameter, is addressed on a continuing basis and as an inherent part of the development and production process. Cost goals are established early in the development phase, which reflect the best balance among life cycle cost, acceptable performance and schedule.

63

Economics (Hitch and McRean) - economics is concerned with allocating resources. Economizing always means trying to make the most efficient use of the resources available. Economics Analysis (Hitch and McRean) - economic analysis - ranging from just straight thinking about alternative courses of action to systematic quantitative comparisons, to identify or achieve near "optimal" solutions. Economic Analysis (DoDI 7041.3) - a systematic approach to the problem of choosing how to employ scarce resources and an investigation of the full implications of achieving a given objective in the most efficient and effective manner. Economic Analysis (NES Dictionary) - a systematic approach to a given problem, designed to assist the manager in solving a problem of choice. The full problem is investigated; objectives and alternatives are searched out and compared in light of their benefits and costs through the use of an appropriate analytical framework. Often used to determine the best use of scarce resources. Effectiveness (DoDI 7041.3) - the performance or output received from an approach or program. Effort Adjustment Factor: EAF - a term used in COCOMO to calculate the cost driver attribute's effect on the project. It is the product of the effort multipliers corresponding to each of the cost drivers for the project. Estimating (NES Dictionary) - to predict costs. Generation of detailed and realistic forecasts of hours, material costs, or other requirements for a task, subtask, operation, or a part or groups thereof - generally in response to a request for proposal or specific statement of work. The art of approximating the probable worth (or cost), extent, quantity, quality, or character of something based on information available at the time. It also covers the generation of predicted costing rates and factors with which estimate numbers may be converted to costs, and the application of these costing rates and factors to "estimate numbers" to establish predicted costs. Fixed Year Dollars - dollars that have been adjusted for inflation to a specific year. Historical Data - a term used to describe a set of data reflecting actual cost or past experience of a product or process. Improvement Curve - a graphical representation of improvement curve theory that projects resource requirements versus the number of units produced. Examples include labor hours, labor cost, direct cost including labor and materials. When reflecting only labor hours the curve is usually called a learning curve. When depicting cost, it is normally referred to as a Cost Improvement Curve. Inflation - an increase in the level of prices for the same item(s). Examples of indices that reflect inflation include Consumer Price Index (CPI), Wholesale Price Index, and Producer Price Index. When prices decline it is called deflation.

64

Normalize - to adjust data (normally cost data) for effects like inflation, anomalies, seasonal patterns, technology changes, accounting system changes, reorganizations, etc. Paradigm - a model, example, or pattern. A generally accepted way of thinking. Parametric Cost Estimate - Estimate derived from statistical correlation of historic system costs with performance and/or physical attributes of the system. Parametric Cost Estimating Methods - Estimating methods based on physical or performance characteristics and schedules of the end items. Parametric Cost Model - A mathematical representation of parametric cost estimating relationships that provides a logical and predictable correlation between the physical or functional characteristics of a system, and the resultant cost of the system. A parametric cost model is an estimating system comprising cost estimating relationships (CERs) and other parametric estimating functions, e.g., cost quantity relationships, inflation factors, staff skills, schedules, etc. Parametric cost models yield product or service costs at designated levels and may provide departmentalized breakdown of generic cost elements. A parametric cost model provides a logical and repeatable relationship between input variables and resultant costs. Parametric Estimating - a mathematical procedure where product or service descriptors (parameters) and cost algorithms directly yield consistent cost information. Process - the sequence of activities (in software development) described in terms of the user roles, user tasks, rules, events, work products, resource use, and the relationships between them. It may include the specific design methodology, language, documentation standards, etc. Rayleigh Distribution - a curve that yields a good approximation to the actual labor curves on software projects. Resource Analysis (Fisher) - systematically determining the economic resource impact of alternative proposals for future courses of action. Reusability - ability to use all or the greater part of the same programming code or system design in another application. Should Cost Estimate - An estimate of contract price which reflects reasonably achievable economy and efficiency. It is generally accomplished by a team of procurement, contract administration, cost analysis, audit and engineering representatives performing an in-depth analysis of cost and cost effects. Its purpose is to develop realistic cost objectives. Technical Non-Cost Data - engineering variables that describe the item, subsystem or system. The source of this data includes engineering documents, engineering drawings and works of a technical nature which are specified to be delivered pursuant to the contract.

65

Top-Down Models - use a method of estimation that estimates the overall cost and effort of the proposed project derived from global properties of the project. The total cost and schedule is partitioned into components for planning purposes. Update - to update an estimate or CER means to utilize the most recent data to make it current, accurate and complete. Validation - in terms of a cost model, a process used to determine whether the model selected for a particular estimate is a reliable predictor of costs for the type of system being estimated

66

APPENDIX II ALL RELATIONSHIPS WITHOUT EXCLUDING OUTLIERS PREDICTING TOTAL MAN HOURS ALL PROJECTS S. NO. TITLE LEANER EQUATION
1 Predicting Total Hrs. Based on Engineer Hrs. 2 Predicting Total Hrs. Based on Tech III Hrs. 3 Predicting Total Hrs. Based on Tech II Hrs. 4 Predicting Total Hrs. Based on Tech I Hrs. 5 Predicting Total Hrs. Based on AutoCAD Hrs. 6 Predicting Total Hrs. Based on Plotter Hrs. 7 Predicting Total Hrs. Based on Clerical Hrs. 8 Predicting Total Hrs. Based on Total Survey Hrs.

LOGARITHMIC R2 EQUATION R2

POLYNOMIAL EQUATION
y = 4.1502x + 584.97

R2
0.563 0.1908 0.8705

y = 7.1315x + 289.97 0.558 y = 288.59Ln(x) - 350.01 0.4281 y = 0.006x2 + 6.1439x + 311.43 y = 4.1502x + 584.97 0.191

y = 4.1198x + 363.44 0.373 y = 189.3Ln(x) - 35.761 0.2916 y = -0.0018x2 + 4.6524x + 337.02 0.3827 y = 3.2241x + 102.24 0.869 y = 388.64Ln(x) - 1207.6 0.6287 y = 1E-04x2 + 3.18x + 104.77 y = 3.4234x + 142.61 0.821 y = 392.1Ln(x) - 1171.1 0.6268 y = -0.0005x2 + 3.6418x + 126.01 0.8242 y = 6.3258x + 264.11 0.607 y = 313.71Ln(x) - 503.56 0.5066 y = -0.0137x2 + 9.4151x + 157.34 0.6422 y = 43.087x + 211.74 0.384 y = 478.14Ln(x) - 388.86 0.3455 y = -0.5506x2 + 62.381x + 86.089 0.4008 y = 2.64x + 369.82 0.509 y = 136.82Ln(x) + 147.13 0.2058 y = 0.0034x2 + 0.8778x + 509.85 0.4445 0.404 0.0992 0.0992

9 Predicting Total Hrs. Based on 3 Man Crew Survey Hrs. y = 12.673x + 482.38 0.467 y = 148.71Ln(x) + 415.76 0.1949 y = 0.0795x2 + 4.2725x + 587.28 0.3461 10 Predicting Total Hrs. Based on 2 Man Crew Survey Hrs. y = 5.2711x + 416.46 0.456 y = 196.03Ln(x) + 28.957 0.2103 y = 0.0082x2 + 3.1599x + 496.29 12 Predicting Total Hrs. Based on Survey Technician Hrs. 13 Predicting Total Hrs. Based on Survey Office Staff Hrs. 14 Predicting Total Hrs. Based on Mileage Cost. 15 Predicting Total Hrs. Based on Printing Cost. 16 Predicting Total Hrs. Based on Expenses. 17 Predicting Total Hrs. Based on Total Prelim Hrs. 18 Predicting Total Hrs. Based on Total Design Hrs. 19 Predicting Total Hrs. Based on Total Construction Hrs. 20 Predicting Total Hrs. Based on Total As Built Hrs. y = 7.2642x + 637.49 0.097 y = 112.79Ln(x) + 524 0.0931 y = -0.1157x2 + 16.43x + 594.27 y = 7.2642x + 637.49 0.097 y = 112.79Ln(x) + 524 0.0931 y = -0.1157x2 + 16.43x + 594.27 11 Predicting Total Hrs. Based on Registered Surveyor Hrs. y = 4.128x + 536.23 0.257 y = 122.56Ln(x) + 366.55 0.1624 y = -0.0001x2 + 3.9509x + 558.53 0.2011

y = 0.2402x + 358.54 0.387 y = 284.13Ln(x) - 1257.3 0.3585 y = -3E-05x2 + 0.4013x + 239.72 0.4223 y = 0.2481x + 472.33 0.293 y = 225.23Ln(x) - 725.41 0.3602 y = -5E-05x2 + 0.5766x + 253.75 0.4853 y = 0.0325x + 654.73 0.115 y = 187.35Ln(x) - 454.94 0.2952 y = -6E-07x2 + 0.06x + 626.75 y = 2.4486x + 595.08 0.103 y = 119.15Ln(x) + 318.15 0.08 y = 0.0007x2 + 2.1441x + 618.32 0.1246 0.076

y = 1.2036x + 159.17 0.738 y = 506.33Ln(x) - 2271.1 0.6525 y = -0.0002x2 + 1.4119x + 110.55 0.7529 y = 1.7057x + 380.19 0.48 y = 313.23Ln(x) - 828.81 0.3823 y = 0.0002x2 + 1.5457x + 410.09 0.4315 y = 3.4617x + 514.23 0.408 y = 177.71Ln(x) + 108.94 0.2157 y = -0.0006x2 + 3.328x + 568.55 0.2451

67

PREDICTING TOTAL PAYMENT ALL PROJECTS S. NO.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

TITLE

LEANER EQUATION R2
0.48 0.236 0.176 0.435 0.425 0.371 0.262 0.336 0.435 0.345 0.122 0.086 0.086 0.361 0.399 0.03 0.155 0.369 0.313 0.299

LOGARITHMIC EQUATION
y = 14097Ln(x) - 17572

POLYNOMIAL R2 EQUATION
y = 0.0597x2 + 218.8x + 26034

R2
0.4919 0.222 0.2084 0.436 0.4257 0.3792 0.262 0.3062 0.3183 0.3289 0.089 0.0794 0.0794 0.3673 0.489 0.1705 0.2052 0.3908 0.2901 0.1703

Predicting Total Payment Based on Engineer Hrs. Predicting Total Payment Based on Tech III Hrs. Predicting Total Payment Based on Tech II Hrs. Predicting Total Payment Based on Tech I Hrs. Predicting Total Payment Based on AutoCAD Hrs. Predicting Total Payment Based on Plotter Hrs. Predicting Total Payment Based on Clerical Hrs. Predicting Total Payment Based on Total Survey Hrs. Predicting Total Payment Based on 3 Man Crew Survey Hrs. Predicting Total Payment Based on 2 Man Crew Survey Hrs. Predicting Total Payment Based on Registered Surveyor Hrs. Predicting Total Payment Based on Survey Technician Hrs. Predicting Total Payment Based on Survey Office Staff Hrs. Predicting Total Payment Based on Mileage Cost. Predicting Total Payment Based on Printing Cost. Predicting Total Payment Based on Expenses. Predicting Total Payment Based on Total Prelim Hrs. Predicting Total Payment Based on Total Design Hrs. Predicting Total Payment Based on Total Construction Hrs. Predicting Total Payment Based on Total As Built Hrs.

y = 382.63x + 11765 y = 244.31x + 24677 y = 150.82x + 20675 y = 125.79x + 9149.3 y = 136.27x + 10630 y = 263.41x + 13969 y = 1968.1x + 9102.2 y = 119.98x + 18375 y = 646.56x + 19677 y = 261.62x + 18874 y = 167.94x + 26173 y = 361.15x + 29089 y = 361.15x + 29089 y = 12.696x + 14379 y = 15.497x + 18243 y = 0.8499x + 31477 y = 194.43x + 23376 y = 47.495x + 11184 y = 71.431x + 19185 y = 152.91x + 24105

0.3442 y = 1.0697x2 + 205.96x + 15936

y = 7881.9Ln(x) + 2875.2 0.1812 y = -0.5193x2 + 288.65x + 15533 y = 14954Ln(x) - 40686 y = 15334Ln(x) - 40139 y = 12264Ln(x) - 14345 y = 23750Ln(x) - 23159 y = 5962.5Ln(x) + 8814.4 y = 8361.4Ln(x) + 14571 y = 5845.6Ln(x) + 15914 y = 5581.9Ln(x) + 23501 y = 5581.9Ln(x) + 23501 y = 13461Ln(x) - 59864 y = 12571Ln(x) - 47090 y = 12276Ln(x) - 44221 y = 3990.2Ln(x) + 21483 y = 20070Ln(x) - 85049 y = 13920Ln(x) - 34981 y = 4683.6Ln(x) + 19136 0.3113 y = -0.0296x2 + 141.69x + 7742.5 0.317 y = -0.0126x2 + 141.28x + 10444 0.2677 y = -0.2999x2 + 333.33x + 11518 0.2297 y = -1.4532x2 + 2021.6x + 8733.5 0.127 y = 0.1546x2 + 39.479x + 24336 0.2017 y = 3.318x2 + 265.38x + 25704 0.1258 y = -0.5269x2 + 250.31x + 26641 0.0795 y = -4.0728x2 + 693.71x + 27307 0.0795 y = -4.0728x2 + 693.71x + 27307 0.2801 y = -0.0008x2 + 16.639x + 11524 0.327 y = -0.0018x2 + 28.226x + 9698.8 0.3396 y = -0.0001x2 + 6.808x + 26265 0.024 y = 2.5173x2 - 242.31x + 31725 0.335 y = -0.0189x2 + 73.351x + 5420.7 0.2547 y = 0.0099x2 + 62.627x + 20470 0.0559 y = 0.1266x2 + 93.779x + 28142

y = 8894.4Ln(x) + 2316.6 0.1363 y = 0.6788x2 + 88.74x + 24634

68

PREDICTING TOTALCOST PER LOT ALL PROJECTS S. NO.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

TITLE

LEANER EQUATION R2
0.0114 0.0004 0.0001

LOGARITHMIC EQUATION R2

POLYNOMIAL EQUATION
y = 0.7461x2 - 120.37x + 5392.6 y = -0.0355x2 + 5.4561x + 2702.2

R2
0.0488 0.0008 0.0027 0.013 0.0139 0.0093 0.0313 0.0098 0.0043 0.0111 0.01 0.0033 0.0033 0.0195 0.0028 0.0031 0.0064 0.0141 0.0249 0.0107

Predicting Cost/ Unit Based on Engineer Hrs. Predicting Cost/ Unit Based on Tech III Hrs. Predicting Cost/ Unit Based on Tech II Hrs. Predicting Cost/ Unit Based on Tech I Hrs. Predicting Cost/ Unit Based on AutoCAD Hrs. Predicting Cost/ Unit Based on Plotter Hrs. Predicting Cost/ Unit Based on Clerical Hrs. Predicting Cost/ Unit Based on Total Survey Hrs. Predicting Cost/ Unit Based on 3 Man Crew Survey Hrs. Predicting Cost/ Unit Based on 2 Man Crew Survey Hrs. Predicting Cost/ Unit Based on Registered Surveyor Hrs. Predicting Cost/ Unit Based on Survey Technician Hrs. Predicting Cost/ Unit Based on Survey Office Staff Hrs. Predicting Cost/ Unit Based on Mileage Cost. Predicting Cost/ Unit Based on Printing Cost. Predicting Cost/ Unit Based on Expenses. Predicting Cost/ Unit Based on Total Prelim Hrs. Predicting Cost/ Unit Based on Total Design Hrs. Predicting Cost/ Unit Based on Total Construction Hrs. Predicting Cost/ Unit Based on Total As Built Hrs.

y = 15.734x + 1285.9 y = -0.9808x + 2483.7 y = 5.1171x + 1769.2 y = -0.5709x + 2288.3 y = -1.9291x + 2507.1 y = -2.7189x + 2370.7

0.0025 y = 121.13Ln(x) + 1720.2 0.0005 y = 0.0487x2 - 16.443x + 3495.7 y = -110.05Ln(x) + 2725 0.0002 y = 0.0284x2 - 20.277x + 4928.8 0.0011 y = -206.45Ln(x) + 3178.2 0.0007 y = 0.0312x2 - 21.737x + 4890.8 0.0005 y = -35.216Ln(x) + 2317.3 0.0003 y = 0.0811x2 - 31.255x + 4183.6

y = -102.29x + 3344.5 0.0094 y = -1064.1Ln(x) + 4608.2 0.0074 y = 10.525x2 - 550.18x + 7000 y = 0.022x2 - 16.83x + 3915.6 0.0137 y = -489.52Ln(x) + 4393.5 0.01 y = -5.9399x + 3129.9 y = -25.11x + 3128.7 y = -8.8729x + 2745.5 y = -2.3427x + 1746.5 y = -32.216x + 3180.3 y = -32.216x + 3180.3 y = -0.4349x + 2808.2 y = 0.0852x + 2102.1 y = -0.043x + 2243.2 y = -4.8773x + 2232.4 y = 1.3552x + 1534.6 0.0057 y = -361.76Ln(x) + 3421.9 0.0035 y = 0.3396x2 - 53.822x + 3195.8 0.0049 y = -269.51Ln(x) + 3172.2 0.0015 y = 0.0835x2 - 32.144x + 3781.2 0.0038 y = -163.73Ln(x) + 2158.9 0.0132 y = 0.091x2 - 32.142x + 3480.1 0.0058 y = -704.91Ln(x) + 4090.2 0.0111 y = 0.4776x2 - 64.616x + 3118.2 0.0058 y = -704.91Ln(x) + 4090.2 0.0111 y = 0.4776x2 - 64.616x + 3118.2 0.0055 y = -363.61Ln(x) + 4692.3 0.0025 y = 0.0003x2 - 2.2616x + 4892.7 0.0002 y = -32.177Ln(x) + 2385.8 0.0003 y = 7E-05x2 - 0.851x + 3369.7 0.0009 y = 22.522Ln(x) + 2042.5 0.0002 y = 5E-06x2 - 0.3086x + 3034.5 0.0108 y = -432.05Ln(x) + 3506.8 0.0276 y = 0.0732x2 - 26.266x + 3300.7 0.0039 y = 185.38Ln(x) + 1070.8 0.0004 y = 0.0073x2 - 10.32x + 5044.3

y = -1.4525x + 2025.2 0.0093 y = -473.03Ln(x) + 4096 0.0233 y = 0.0165x2 - 17.406x + 4699 y = 0.0416x2 - 22.809x + 3324.3 0.008 y = -115.31Ln(x) + 2183.2 0.0021 y = -3.687x + 2032

69

PREDICTING TOTAL MAN HOURS APARTMENTS S. NO.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

TITLE

LEANER EQUATION R2
0.5619 0.2326 0.2785 0.9191 0.9059 0.7173 0.4049 0.1444

LOGARITHMIC EQUATION
y = 335.55Ln(x) - 574.58 y = 205.9Ln(x) - 73.48 y = 351.85Ln(x) - 1085.3 y = 296.89Ln(x) - 790.6 y = 311.38Ln(x) - 620.33 y = 471.26Ln(x) - 406.34 y = 27.615Ln(x) + 479.3

POLYNOMIAL R2 EQUATION
y = -0.0054x2 + 4.7124x + 488.39

R2
0.5628 0.25 0.3107 0.9194 0.9059 0.7369 0.4463 0.1682 0.1503 0.1981 0.0323 0.0254 0.3763 0.5982 0.3163 0.0289 0.8134 0.5027 0.0964

Predicting Total Hrs Based on Engineer Hrs. Predicting Total Hrs Based on Tech III Hrs. Predicting Total Hrs Based on Tech II Hrs. Predicting Total Hrs Based on Tech I Hrs. Predicting Total Hrs Based on AutoCAD Hrs. Predicting Total Hrs Based on Plotter Hrs. Predicting Total Hrs Based on Clerical Hrs. Predicting Total Hrs Based on Total Survey Hrs. Predicting Total Hrs Based on 3 Man Crew Survey Hrs. Predicting Total Hrs Based on 2 Man Crew Survey Hrs. Predicting Total Hrs Based on Registered Surveyor Hrs. Predicting Total Hrs Based on Survey Technician Hrs. Predicting Total Hrs Based on Survey Office Staff Hrs. Predicting Total Hrs Based on Mileage Cost. Predicting Total Hrs Based on Printing Cost. Predicting Total Hrs Based on Expenses. Predicting Total Hrs Based on Total Prelim Hrs. Predicting Total Hrs Based on Total Design Hrs. Predicting Total Hrs Based on Total Construction Hrs. Predicting Total Hrs Based on Total As Built Hrs.

y = 7.9966x + 219.67 y = 2.8161x + 533.57 y = 4.9372x + 379.8 y = 2.7863x + 101.07 y = 2.9661x + 103.34 y = 6.3875x + 125.56 y = 35.753x + 237.96 y = 2.6753x + 414.88 y = 16.385x + 311.16 y = 5.7271x + 415.88 y = -4.446x + 655.82 y = 15.829x + 478.93 y = 15.829x + 478.93 y = 0.1669x + 381.23 y = 0.1621x + 387.34 y = 0.0852x + 523.92 y = -0.1481x + 642.49 y = 1.1367x + 122.47 y = 1.6952x + 393 y = 1.8247x + 631.86

0.5261 y = -0.0063x2 + 8.9108x + 199.93 0.3057 y = -0.0242x2 + 8.7285x + 298.85 0.7133 y = 0.0003x2 + 2.6227x + 113.1 0.6418 y = 2E-05x2 + 2.9582x + 103.9 0.596 y = -0.0142x2 + 9.2211x + 36.947 0.4197 y = -0.7437x2 + 67.33x + 35.628 0.0084 y = 0.031x2 - 3.7949x + 615.81

0.6891 y = 164.02Ln(x) + 251.19 0.4256 y = 0.1801x2 - 0.1738x + 559.03 0.2185 y = 59.263Ln(x) + 431.76 0.0262 y = 0.081x2 - 3.9001x + 572.17 0.0517 y = -21.025Ln(x) + 609.39 0.0048 y = 0.013x2 - 4.9173x + 645.71 0.0857 y = 47.225Ln(x) + 525.02 0.0239 y = 0.1443x2 + 7.0487x + 562.79 0.0857 y = 47.225Ln(x) + 525.02 0.0239 0.3266 0.3989 0.1047 0.8015 0.4317 y = 290.03Ln(x) - 1359.3 y = 246.16Ln(x) - 1024.4 y = 166.98Ln(x) - 375.54 y = 489.83Ln(x) - 2233.8 y = 255.32Ln(x) - 519.69 0.3934 y = -3E-05x2 + 0.3585x + 245.4 0.4601 y = -3E-05x2 + 0.4156x + 189.75 0.2906 y = -4E-05x2 + 0.3694x + 394.47 0.052 y = 0.0128x2 - 3.5876x + 628.41 0.7354 y = -0.0003x2 + 1.49x + 45.294 0.4035 y = -0.0038x2 + 3.6882x + 238.62

0.0012 y = -71.464Ln(x) + 882.24

0.0977 y = 31.144Ln(x) + 627.08 0.0161 y = -0.0114x2 + 4.815x + 525.9

70

PREDICTING TOTAL PAYMENT APARTMENTS S. NO.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

TITLE

LEANER EQUATION R2
0.5789 0.3153 0.1315 0.5564 0.5328 0.5006 0.4273 0.2108 0.8708 0.2844

LOGARITHMIC EQUATION
y = 22389Ln(x) - 42386 y = 15176Ln(x) - 13930 y = 18104Ln(x) - 50734 y = 15193Ln(x) - 35176 y = 16388Ln(x) - 28243 y = 29249Ln(x) - 26475 y = 3359.3Ln(x) + 21567 y = 14371Ln(x) + 5349.1 y = 6899.2Ln(x) + 12878

POLYNOMIAL R2
0.308

EQUATION
y = -0.3617x2 + 443.49x + 17147 y = 0.4806x2 - 2.136x + 33216

R2
0.588 0.324 0.1402 0.574 0.5423 0.5171 0.4289 0.2663 0.2446 0.2891 0.0416 0.0633 0.055 0.3861 0.6635 0.396 0.0899 0.6343 0.2323 0.0534

Predicting Total Payment Based on Engineer Hrs. Predicting Total Payment Based on Tech III Hrs. Predicting Total Payment Based on Tech II Hrs. Predicting Total Payment Based on Tech I Hrs. Predicting Total Payment Based on AutoCAD Hrs. Predicting Total Payment Based on Plotter Hrs. Predicting Total Payment Based on Clerical Hrs. Predicting Total Payment Based on Total Survey Hrs. Predicting Total Payment Based on 3 Man Crew Survey Hrs. Predicting Total Payment Based on 2 Man Crew Survey Hrs. Predicting Total Payment Based on Registered Surveyor Hrs. Predicting Total Payment Based on Survey Technician Hrs. Predicting Total Payment Based on Survey Office Staff Hrs. Predicting Total Payment Based on Mileage Cost. Predicting Total Payment Based on Printing Cost. Predicting Total Payment Based on Expenses. Predicting Total Payment Based on Total Prelim Hrs. Predicting Total Payment Based on Total Design Hrs. Predicting Total Payment Based on Total Construction Hrs. Predicting Total Payment Based on Total As Built Hrs.

y = 590.87x + 7883.9 y = 385.69x + 18462 y = 156.47x + 31256 y = 157.81x + 7725.7 y = 165.59x + 8256.2 y = 388.45x + 7131.7 y = 2673.7x + 8912.4 y = 238.53x + 18962 y = 1479x + 9939.6 y = 498.7x + 16789 y = -298.15x + 38049 y = 2023.1x + 18745 y = 764.47x + 33538 y = 11.47x + 20949 y = 13.902x + 17754 y = 8.2025x + 28667 y = 87.413x + 36778 y = 71.808x + 5740.1 y = 75.652x + 26727 y = 61.501x + 41890

0.442 y = 1.4258x2 + 382.44x + 12386

0.3564 y = 0.1832x2 + 66.22x + 14468 0.3172 y = 0.147x2 + 95.906x + 13163 0.3115 y = 0.9475x2 + 199.33x + 13046 0.3051 y = 10.509x2 + 2227.5x + 11772 0.0227 y = 2.9268x2 - 375.9x + 38267 0.5068 y = 18.934x2 - 177.14x + 32703

0.061 y = 8.6897x2 - 587.59x + 37140 1.00E0.0454 y = -7.1742Ln(x) + 31549 07 y = -4.873x2 - 44.568x + 38558 0.2842 0.0332 0.2911 0.5534 0.183 0.5967 0.1577 0.021 y = 8795.8Ln(x) + 21068 y = 11325Ln(x) + 29620 y = 20291Ln(x) - 101092 y = 18426Ln(x) - 85609 y = 11624Ln(x) - 31926 y = 26953Ln(x) - 119942 y = 12563Ln(x) - 19586 0.1685 y = 29.255x2 + 468.45x + 32420 0.1808 y = -161.53x2 + 5843.2x + 29204 0.3634 y = -0.0031x2 + 30.735x + 7291.8 0.4865 y = -0.0017x2 + 27.618x + 7062.6 0.2657 y = -0.0027x2 + 28.95x + 19214 0.4154 y = 0.0401x2 + 16.291x + 17683 0.1792 y = -0.312x2 + 234.19x + 16166

0.0667 y = -1507.4Ln(x) + 48528 0.0038 y = 1.1925x2 - 237.7x + 36401

y = -1528.8Ln(x) + 49326 0.0073 y = -0.9173x2 + 315.29x + 31661

71

72

PREDICTING TOTAL PAYMENT MISCELLANEOUS PREDICTING TOTAL MAN HOURS MISCELLANEOUS S. NO.
1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20

TITLE TITLE

LEANER LEANER EQUATION EQUATION

2 R R2

LOGARITHMIC LOGARITHMIC EQUATION EQUATION

y = 254.82Ln(x) - 191.95
2 R R2

POLYNOMIAL POLYNOMIAL EQUATION EQUATION

2 R R2

Predicting Based on Engineer Predicting Total Total Payment Hrs Based on Engineer Hrs. Hrs. Predicting Based on Tech III Hrs. Predicting Total Total Payment Hrs Based on Tech III Hrs. Predicting Based on Tech II Hrs. Predicting Total Total Payment Hrs Based on Tech II Hrs. Predicting Based on Tech Predicting Total Total Payment Hrs Based on Tech I Hrs.I Hrs. Predicting Based on AutoCAD Predicting Total Total Payment Hrs Based on AutoCAD Hrs. Hrs. Predicting Based on Plotter Predicting Total Total Payment Hrs Based on Plotter Hrs. Hrs. Predicting Based on Clerical Predicting Total Total Payment Hrs Based on Clerical Hrs. Hrs. Predicting Total Total Hrs Payment Based on Total Survey Predicting Based on Total Survey Hrs. Hrs. Predicting Based on 3 Crew Man Crew Survey Predicting Total Total Payment Hrs Based on 3 Man Survey Hrs. Hrs. Predicting Based on 2 Crew Man Crew Survey Predicting Total Total Payment Hrs Based on 2 Man Survey Hrs. Hrs. Predicting Based on Registered Surveyor Predicting Total Total Payment Hrs Based on Registered Surveyor Hrs. Hrs. Predicting Based on Survey Technician Predicting Total Total Payment Hrs Based on Survey Technician Hrs. Hrs. Predicting Based on Survey Office Staff Hrs. Predicting Total Total Payment Hrs Based on Survey Office Staff Hrs. Predicting Based on Mileage Cost. Cost. Predicting Total Total Hrs Payment Based on Mileage Predicting Based on Printing Cost. Cost. Predicting Total Total Hrs Payment Based on Printing Predicting Based on Expenses. Predicting Total Total Hrs Payment Based on Expenses. Predicting Total Hrs Based on Total Prelim Hrs. Hrs. Predicting Total Payment Based on Total Prelim Predicting Based on Total Design Hrs. Hrs. Predicting Total Total Hrs Payment Based on Total Design Predicting Based on Total Construction Hrs. Hrs. Predicting Total Total Hrs Payment Based on Total Construction Predicting Total Total Payment Hrs Based on Total As Built Predicting Based on Total As Hrs. Built Hrs.

y y= = 160.22x 8.2822x + + 17488 205.3 y = 347.95x + 12703 y = 15.753x + 238.33 y= = 4.9637x 117.25x + + 245.34 12745 y y= = 3.2856x 54.867x + + 42.585 11958 y y= = 3.5355x 57.293x + + 73.932 12752 y y= = 8.1487x 166.66x + + 119.86 11368 y y= = 51.995x 573.39x + + 48.799 15305 y y= = 4.0808x 91.334x + + 259.38 14639 y y= = 17.339x 436.75x + + 384.18 16988 y y y= = 230.95x 10.632x + + 13492 175.5 y= = 10.929x 241.28x + + 379.67 15698 y y= = 19.23x 338.8x + + 317.71 19064 y

0.1003 0.7688 0.187 0.4733

0.5045 y y= = 12358e0.0057x 0.0082x2 + 6.8608x + 232.61 y = 296.87x++250.36 24277 y = -3.3003x2 -0.042x2 ++18.052x 0.185 0.3314 0.317 0.7283 0.2957 0.7017 0.2881 0.5455 y y= = 0.2003x2 0.0017x2 +13.258x 4.3973x + +25878 267 y = -0.0255x2 + 24.592x + 24564 y = 0.0011x2 + 2.7068x + 86.511 y ++ 32.71x + 24554 y= = -0.05x2 0.0018x2 2.7018x + 126.67 y = 0.937x2 117.38x + y = -0.0374x2 + 14.246x28412 - 21.679

0.0958 0.7711 0.0077 0.4465 0.0296 0.6768 0.0043 0.9087 0.0037 0.8563 0.0083 0.6695 0.0573 0.5974 0.0278 0.5326 0.0507 0.4826 0.0285 0.5442 0.0113 0.4117 0.007 0.2544 0.0398 0.2544 0.6829 0.058 0.7744 0.0243 0.8161 0.0197 0.0644 0.0327 0.7213 0.0046 0.3514 0.0676 0.4213 0.0414

0.2978 4627.2Ln(x)++13.585 5481 0.6751 yy==174.16Ln(x) 0.1997 y= =454.41Ln(x) 10662Ln(x) - 1552.1 29370 0.9057 y 0.1769 y= = 431.11Ln(x) 9952.4Ln(x) -- 1373.6 24427 0.8519 y

0.2123 y= = 336.06Ln(x) 8684.7Ln(x) -- 611.12 10222 0.6419 y 0.0573 3992.6 0.5051 0.058 y ++ 15305 0.5963 y y= = 7901.5Ln(x) 655.42Ln(x) + - 840.82 y= = 573.39x 0.2556x2 42.582x + 107.02 0.2522 y y= =169.73Ln(x) 5154Ln(x) +-3783.8 y= 0.6851 9.9127 0.1851 0.2732 y = 0.1414x2 0.0078x2 +9.0894x 0.4788x + +25981 474.54 0.2921 y = 8852.1Ln(x) + 7801.2 0.3456 y = -1.7194x2 + 469.17x + 23807 0.6362 y = 272.84Ln(x) + 158.34 0.4537 y = 0.0966x2 + 7.3624x + 489.54 0.255 y 0.7378 y= = 14202Ln(x) 452.01Ln(x)--24734 900.3 0.2464 y = 5095.8Ln(x) + 10547 0.3743 y = 160.04Ln(x) + 330.18 0.087 y y 5248.9Ln(x) 13601 0.3847 == 219.52Ln(x) ++ 168.75 0.3653 0.5051 0.0304 0.2217 y +3.0051x 134.64x+ +430.27 23397 y= = -0.0505x2 0.0349x2 + y y= = 3.1066x2 0.2506x2 - 164.91x 10.383x + + 26765 530.41 y y= = 3.0213x2 0.2718x2 +5.6516x 1.2822x + +26294 506.18 y = 319.52x2 1941x + 26995 y = 0.2718x2 + 1.2822x + 506.18 y y= = 8E-06x2 0.0042x2+- 0.3818x 12.707x + + 229.93 30906 y = 0.0001x2 + 0.3902x + 197.0 y = 0.0059x2 - 5.869x + 26025 y y= = -1E-05x2 -1E-04x2 + + 0.6246x 5.6817x + + 255.69 24317 y = 0.0591x2 4.4312x + 599.79 y = 3.8303x2 - 385.11x + 28655 y y= = 0.0002x2 0.0029x2 + + 0.8629x 0.8533x + + 204.51 26053 y = -8E-05x2 + 1.3718x + 406.97 y = 0.0983x2 - 62.64x + 29322

0.1202 0.2885 yy == 319.52x2 1941x + 0.0398 y = 5844.6Ln(x) + 25545 19.23x + 317.71 0.3847 y = 219.52Ln(x) + 168.75 0.0713 0.2885 26995 y y= = 0.4148x 8.9435x + + 213.61 12834 0.6824 0.2508 y y= = 276.22Ln(x) 6007.1Ln(x) -- 1141.7 16715 0.3761 0.1406 y = 0.674x + 124.79 0.7582 y = 252.12Ln(x) 871.44 y = 15.86x + 9942.4 0.3319 y = 6905.3Ln(x) - 19256 0.5247 0.3112 y y= = 0.0356x 0.4653x + + 540.47 20564 yy==4.8309x + 327.29 193.9x + 13081 y= = 19.138x 1.0868x + + 13561 163.5 y y y= = 1.6497x 42.078x + + 233.63 11633 y y= = 8.2225x 119.16x + + 216.43 11620 0.4612 0.0623 y y= = 322.33Ln(x) 6954.9Ln(x) -- 1268.7 19155 0.3723 y = 166.84Ln(x) + 9.8834 0.3703 y = 6494.5Ln(x) + 1094.1 0.7182 y= = 9283.9Ln(x) 409.2Ln(x) - -1675.8 0.1761 y 30492 0.6255 0.5401 0.8303 0.6457 y y= = 320.81Ln(x) 8120.2Ln(x) -- 948.98 18227 yy==550.52Ln(x) 7461.9Ln(x)- -1250.5 7862 0.7984 0.2939 0.2542 0.2378 0.5983 0.2435 0.493 0.4192

0.6043 y= = 2.4015x2 0.051x2 - 6.2809x ++ 641.32 0.411 y 582.41x 34744

73

PREDICTING TOTAL MAN HOURS SUBDIVISION S. NO.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

TITLE

LEANER EQUATION R2
0.4638 0.207 0.2749 0.8711 0.8368 0.6095 0.2837 0.5375

LOGARITHMIC EQUATION
y = 294.5Ln(x) - 360.86 y = 199.02Ln(x) - 84.584 y = 376.22Ln(x) - 1110.4 y = 456.78Ln(x) - 1448.8 y = 306.65Ln(x) - 408.16 y = 403.22Ln(x) - 178.62 y = 142.67Ln(x) + 154.1

POLYNOMIAL R2 EQUATION
y = 0.0333x2 - 1.0212x + 724.35

R2
0.4808 0.228 0.3087 0.8744 0.8439 0.6556 0.3003 0.5692 0.3782 0.4925 0.2835 0.0916

Predicting Total Hrs Based on Engineer Hrs. Predicting Total Hrs Based on Tech III Hrs. Predicting Total Hrs Based on Tech II Hrs. Predicting Total Hrs Based on Tech I Hrs. Predicting Total Hrs Based on AutoCAD Hrs. Predicting Total Hrs Based on Plotter Hrs. Predicting Total Hrs Based on Clerical Hrs. Predicting Total Hrs Based on Total Survey Hrs. Predicting Total Hrs Based on 3 Man Crew Survey Hrs. Predicting Total Hrs Based on 2 Man Crew Survey Hrs. Predicting Total Hrs Based on Registered Surveyor Hrs. Predicting Total Hrs Based on Survey Technician Hrs. Predicting Total Hrs Based on Survey Office Staff Hrs. Predicting Total Hrs Based on Mileage Cost. Predicting Total Hrs Based on Printing Cost. Predicting Total Hrs Based on Expenses. Predicting Total Hrs Based on Total Prelim Hrs. Predicting Total Hrs Based on Total Design Hrs. Predicting Total Hrs Based on Total Construction Hrs. Predicting Total Hrs Based on Total As Built Hrs.

y = 6.2485x + 369.41 y = 5.2213x + 633.24 y = 3.6818x + 400.29 y = 3.4401x + 113.9 y = 3.6898x + 170.66 y = 5.9166x + 366.67 y = 41.94x + 269 y = 2.5058x + 371.35 y = 10.021x + 598.92 y = 4.5596x + 472.66 y = 3.9112x + 557.57 y = 3.1657x + 822.24 y = 3.1657x + 822.24 y = 0.2476x + 372.46 y = 0.5858x + 300.74 y = 0.0162x + 751.04 y = 3.7899x + 559.73 y = 1.2802x + 167.97 y = 1.7494x + 405.19 y = 3.3101x + 522.28

0.341 y = 0.0139x2 + 3.8555x + 429.31 0.2323 y = -0.0127x2 + 6.982x + 243.4 0.5531 y = 0.0001x2 + 3.3866x + 116.32 0.6233 y = -0.0005x2 + 3.9441x + 147.74 0.4829 y = -0.0131x2 + 9.2356x + 238.07 0.2439 y = -0.2749x2 + 50.595x + 210.08 0.2027 y = 0.003x2 + 0.9026x + 506.05

0.3508 y = 74.533Ln(x) + 669.89 0.0476 y = 0.0875x2 + 1.7083x + 656.77 0.4422 y = 189.05Ln(x) + 80.829 0.1768 y = 0.0082x2 + 2.6933x + 517.98 0.3176 y = 140.96Ln(x) + 315.57 0.1914 y = -0.0012x2 + 4.1148x + 558.08 0.0289 y = 54.898Ln(x) + 756.52 0.0197 y = -0.1233x2 + 15.89x + 664.29 y = -0.1233x2 + 15.89x + 0.0289 664.29 0.0916 0.3454 0.4968 0.0117 0.7298 0.4606 0.4464 y = 292.37Ln(x) - 1307.7 y = 215.57Ln(x) - 591.92 y = 131.88Ln(x) - 73.128 y = 613.95Ln(x) - 2911.2 y = 292.08Ln(x) - 377.85 y = 292.08Ln(x) - 377.85 0.3137 y = -1E-05x2 + 0.308x + 320.56 0.2976 y = -3E-05x2 + 0.6637x + 260.67 0.1333 y = -4E-06x2 + 0.1001x + 669.43 0.6604 y = -0.0001x2 + 1.5028x + 102.41 0.3295 y = 0.0012x2 + 0.8438x + 505.88 0.3724 y = -0.0004x2 + 2.9921x + 602.51

0.3595 0.5189 0.0643 0.2036 0.7569 0.4688 0.2803

0.1799 y = 164.03Ln(x) + 199.31 0.1277 y = 0.0258x2 - 1.1257x + 688.36

74

PREDICTING TOTAL PAYMENT SUBDIVISION S. NO. S. NO. 1

2 3 1 4 2 5 3 6 4 7 5 8 6 9 7 10 8 9 11 10 12 11 13 12 14 13 15 14 16 15 17 16 18 17 19 18 20 19 20

TITLE LEANER LOGARITHMIC PREDICTING TOTAL MAN HOURS WITHOUT OUTLIERS ALL PROJECTS TITLE
Predicting Total Payment Based on Engineer Hrs. Predicting Total Payment Based on Tech III Hrs. Predicting Total Total Hrs. Payment Based on TechHrs. II Hrs. Predicting Based on Engineer Payment Based on Tech I Hrs. Predicting Total Hrs. Based on Tech III Hrs. Payment Based on AutoCAD Predicting Total Hrs. Based on Tech II Hrs. Hrs. Payment Based on Plotter Predicting Total Hrs. Based on Tech I Hrs. Hrs. Payment Based on Clerical Hrs. Predicting Total Hrs. Based on AutoCAD Hrs. Predicting Total Payment Based on Total Survey Hrs. Hrs. Based on Plotter Hrs. Predicting Total Payment Based on 3 Man Crew Survey Hrs. Hrs. Based on Clerical Hrs. Predicting Total Payment Based on 2Survey Man Crew Hrs. Based on Total Hrs. Survey Hrs. Predicting Total Hrs. Based on 3 Man Crew Survey Hrs. Predicting Total Payment Based on Registered Surveyor Hrs. Predicting Total Hrs. Based on 2 Man Crew Survey Hrs. Predicting Total Payment Based on Survey Technician Hrs. Predicting Total Hrs. Based on Registered Surveyor Hrs. Predicting Total Payment Based on Survey Office Staff Hrs. Predicting Total Hrs. Based on Survey Technician Hrs. Predicting Total Payment Based on Mileage Cost. Predicting Total Hrs. Based on Survey Office Staff Hrs. Predicting Total Payment Based on Printing Cost. Predicting Total Hrs. Based on Mileage Cost. Predicting Total Payment Based on Expenses. Predicting Total Hrs. Based on Printing Cost. Predicting Total Payment Based on Total Prelim Hrs. Predicting Total Hrs. Based on Expenses. Predicting Total Payment Based on Total Design Hrs. Predicting Total Hrs. Based on Total Prelim Hrs. Predicting Total Payment Based on Total Construction Hrs. Predicting Total Hrs. Based on Total Design Hrs. Predicting Total Payment Based on Total As Built Hrs. Predicting Total Hrs. Based on Total Construction Hrs. Predicting Total Hrs. Based on Total As Built Hrs.

POLYNOMIAL EQUATION POLYNOMIAL R2

0.1846

EQUATION LEANER
y = 321.52x + 15274

R2
0.4233

EQUATION LOGARITHMIC R2
y = 14503Ln(x) - 19981

0.3034 y = 17803e0.0074x

EQUATION y= 318.89x + 27756

y= = 6.6607x 165.19x + + 297.46 18851 y y= + 8122 y 144.52x = 4.8853x + 481.98 158.98x + 288.53 9806.9 y = 4.0078x y= = 3.1213x 236.93x + + 104.99 19508 y y= = 3.2036x 1761.8x + + 158.29 14360 y y= = 6.9456x 113.69x + + 204.89 16894 y y= = 41.995x 526.63x + + 169.23 25539 y yy == 223.93x ++ 20001 2.6057x 287 y = 14.071x + 380.96 y = 126.95x + 28860 y = 5.7334x + 360.85 y = 271.43x + 35277 y = 4.2487x + 472.61 y = 989.77x + 33361 y = 17.23x + 470.31 y = 11.891x + 16230 y = 26.895x + 594.07 y = 27.701x + 13382 y = 0.2168x + 326.93 y = 1.297x + 33839 y = 0.4636x + 259 y = 242.29x + 21764 y = 0.2751x + 471.78 y = 48.993x + 12561 y = 2.1479x + 560.15 y = 88.552x + 16616 y = 1.1862x + 132.55 y = 166.23x + 22007 y = 1.6489x + 348.87 y = 3.0771x + 465.99

R2 0.2847
0.2062 0.6335 0.5553 0.1377 0.5463 0.5686 0.3576 0.9097 0.184 0.8505 0.3785 0.7622 0.3337 0.4871 0.3543 0.6724 0.7186 0.1054 0.6117 0.0684 0.4677 0.0895 0.3118 0.298 0.2232 0.4093 0.4356 0.0283 0.7064 0.2714 0.3145 0.3756 0.0871 0.4206 0.8379 0.3862 0.5732 0.5526

EQUATION
y 2075.1 y= =8748.1Ln(x) 267.6Ln(x) --295.26 y = 15356Ln(x) - 41165 y= =159.06Ln(x) 19190Ln(x) - 2.9569 57688 y y= =346.71Ln(x) 11900Ln(x) - 1037.8 10055 y y= = 349.47Ln(x) 18407Ln(x) y - 7956.3 999.32 y = 284.14Ln(x) 5537.6Ln(x) +434.33 11542 y = 405.85Ln(x) 3890.7Ln(x) +286.96 29339 y= = 140.6Ln(x) 7473Ln(x) + y + 8124.5 29.366

R2

EQUATION y = 2.5707x2 - 163.78x + 34866

R 0.342
0.1448 0.6377 0.4367 0.0776 0.4319 0.5808 0.2986 0.9113 0.1219 0.8555 0.4386 0.7716 0.3428 0.4945 0.4442 0.4223 0.3917 0.1007 0.477 0.083 0.2311 0.112 0.1787 0.2521 0.2431 0.3385 0.4487 0.0673 0.7297 0.3191 0.3271 0.3166 0.0399 0.38 0.8489 0.2142 0.4282 0.1897

0.1683 y = -0.0052x2 -0.0982x2 + + 7.5716x 160.81x + + 272.53 21863 0.5085 y= 0.3417 y = -0.0273x2 0.1271x2 + +57.899x 6.1156x+ +19494 517.27 0.3975 y = 0.0029x2 0.1055x2 + 3.3135x 87.197x + 308.79 18855 0.3663 0.2684 y = -0.0005x2 -0.3264x2 + 3.3637x 294.54x + 84.011 19329 0.6592 0.1828 y = -0.0012x2 12.502x2 + 0.6743 +1062.8x 3.7263x+ +21671 120.05 0.1031 29.876x 0.6033 y = 0.2643x2 -0.0089x2-+ 8.536x + 29859 158.37 0.0446 49.359x ++ 32087 0.4073 y = 5.4991x2 -0.1037x2-+ 45.452x 145.99 y = 1.0987x2 0.0037x2 +60.098x 0.7941x + +30695 452.83 y = 0.0704x2 + 5.1938x + 544.45 y = -0.3204x2 + 180.01x + 28952 y = 0.0053x2 + 3.9975x + 454.53 y = -3.1565x2 + 586.48x + 31788 0.0023x2 + 3.2638x + 539.65 y = -36.365x2 + 2176.2x + 29405 y = 0.2442x2 + 2.2631x + 595.92 y = 0.0028x2 - 0.9843x + 27651 y = -0.5557x2 + 43.745x + 543.06 y = 0.007x2 + 9.5463x + 22603 y = -2E-05x2 + 0.2924x + 277.01 y = -0.0002x2 + 5.2428x + 31121 y = -6E-05x2 + 0.6405x + 184.07 y = 2.302x2 - 203.24x + 33857 y = -3E-05x2 + 0.3885x + 426.01 y = -0.0086x2 + 54.944x + 13567 y = -0.0039x2 + 2.2615x + 604.05 y = 0.0933x2 + 10.49x + 27574 y = -0.0002x2 + 1.4015x + 87.763 y = 0.2071x2 + 54.531x + 31238 y = 0.0008x2 + 1.0941x + 415.98

0.0849 0.3648 5.94Ey = 197.9Ln(x) + 218.78 0.4148 y = 4455Ln(x) + 21462 02 y = 254.13Ln(x) - 222.21 0.3742 y = 3814.6Ln(x) + 31705 0.0306 y = 124.49Ln(x) + 297.81 0.3014 y = 143.32Ln(x) + 416.42 y = 12871Ln(x) - 56049 y = 161.93Ln(x) + 549.13 y = 9230.3Ln(x) - 22725 y = 218.08Ln(x) - 877.6 y = 7190.1Ln(x) - 10685 y = 214.52Ln(x) - 702.1 y = 8846.4Ln(x) + 5015.8 y = 215.45Ln(x) - 621.85 y = 24706Ln(x) - 112669 y = 94.728Ln(x) + 347.78 y = 17323Ln(x) - 53090 y = 448.83Ln(x) - 1983.1 y = 13496Ln(x) - 18429 y = 255.26Ln(x) - 594.84 y = 204.76Ln(x) - 82.671 0.1621 0.2283 0.2583 0.2028 0.3866 0.1456 0.5032 0.1376 0.3523 0.3724 0.0785 0.2817 0.749 0.2633 0.4473

0.4539 y = 0.0001x2 + 2.485x + 557.18

APPENDIX III ALL RELATIONSHIPS AFTER EXCLUDING OUTLIERS

75

PREDICTING PREDICTING TOTAL TOTAL MAN PAYMENT HOURS WITHOUT WITHOUT OUTLIERS OUTLIERS ALL APARTMENTS APARTMENTS PREDICTING TOTAL PAYMENT WITHOUT OUTLIERS PROJECTS S. NO.
1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20

TITLE TITLE

LEANER LEANER EQUATION EQUATION

2 R R2

LOGARITHMIC LOGARITHMIC EQUATION EQUATION

y y= = 13696Ln(x) 17328 y =355.11Ln(x) 12378Ln(x) - 639.35 12698
2 R R2

POLYNOMIAL POLYNOMIAL EQUATION EQUATION

2 R R2

Predicting Hrs Based Based on Engineer on Hrs. Hrs. Predicting Total Total Payment Payment Based on Engineer Engineer Hrs. Predicting Total Hrs Payment Based Based on Tech on Tech III Hrs. III Hrs. Predicting Total Payment Based on Tech III Hrs. Predicting Hrs Based Based on Tech on II Hrs. II Predicting Total Total Payment Payment Based on Tech Tech II Hrs. Hrs. Predicting Total Hrs Payment Based Based on Tech on Tech I Hrs. I Hrs. Predicting Total Payment without MEP Based on Tech I Hrs. Predicting Hrs Based without on AutoCAD Based Hrs. on AutoCAD Hrs. Predicting Total Total Payment Payment Based on AutoCAD Hrs. Predicting Hrs Based Based on Plotter on Hrs. Hrs. Predicting Total Total Payment Payment Based on Plotter Plotter Hrs. Predicting Hrs Based Based on Clerical on Hrs. Hrs. Predicting Total Total Payment Payment Based on Clerical Clerical Hrs. Predicting Total Hrs Payment Based Based on Total on Total Survey Survey Hrs. Hrs. Hrs. Predicting Total Payment without MEP Based on Total Survey Predicting Hrs Based Based on 3 Man on Man Survey Survey Hrs. Hrs. Predicting Total Total Payment Payment Based on 3 3 Crew Man Crew Crew Survey Hrs. Predicting Total Hrs Payment Based Based on 2 Man on 2 Crew Man Crew Survey Survey Hrs. Predicting Total Payment Based on 2 Man Crew Survey Hrs. Hrs. Predicting Hrs Based Based on Registered on Surveyor Surveyor Hrs. Hrs. Predicting Total Total Payment Payment Based on Registered Registered Surveyor Hrs. Predicting Total Total Payment Hrs Based Payment Based on Survey on Survey Survey Technician Technician Hrs. Hrs. Hrs. Predicting Based on Technician Predicting Hrs Based Based on Survey on Office Office Staff Hrs. Staff Predicting Total Total Payment Payment Based on Survey Survey Office Staff Hrs. Hrs. Predicting Total Hrs Payment Based Based on Mileage on Mileage Cost. Cost. Predicting Total Payment Based on Mileage Cost. Predicting Hrs Based Based on Printing on Cost. Cost. Predicting Total Total Payment Payment Based on Printing Printing Cost. Predicting Total Hrs Payment Based Based on Expenses. on Expenses. Predicting Total Payment Based on Expenses. Predicting Hrs Based Based on Total on Prelim Prelim Hrs. Hrs. Predicting Total Total Payment Payment Based on Total Total Prelim Hrs. Predicting Total Hrs Payment Based Based on Total on Total Design Design Hrs. Hrs. Hrs. Predicting Total Payment without MEP Based on Total Design Predicting Hrs Based Based on Total on Construction Construction Hrs. Hrs. Predicting Total Total Payment Payment Based on Total Total Construction Hrs. Predicting Total Hrs Payment Based Based on Total on Total As Built As Hrs. Built Hrs. Predicting Total Payment Based on Total As Built Hrs.

y y= = 372.76x + 13777 y = 8.6459x 317.17x + + 186.59 14776 y y = = 4.6306x 304.29x + + 384.26 y = 294.82x + 19922 21394 y y = 216.25x + 16762 y= = 4.5986x 172.99x + + 273.75 15775 y y = = 127.65x 2.955x + + 85.256 7048.8 y = 138.01x + 7367.6 y y= =134.93x 3.1146x+ +7199.1 93.24 y = 131.98x + 11561 y y= =7.4265x 294.63x+ +72.492 5856 y = 286.96x + 11924 y 45.023x 103.09 y= = 2734.9x 2125.6x + + 4755.7 8106.1 y y = = 2.5897x 166.14x + + 16329 y = 147.65x + 259.65 13268 y y = 938.27x + 13997 y= = 16.211x 772.48x + + 278.09 15153 y y = = 7.4499x 365.6x + + 14278 316.82 y = 311.75x + 16892 y y= =-20.193x 1.9897x + + 23236 347.4 21984 y = 174.77x y y = 942.18x + +26249 459.04 20054 y= =2.8962x 313.1x + y y = 775.39x + 24473 y= = 91.112x 1252.2x + + 249.28 29565 y y = = 0.2818x 11.859x + + 207.39 y = 10.338x + 15871 14936 y 0.3988x 179.09 y= = 20.439x 20.719x + + 7774.3 12154 y y = = 12.381x 0.2119x + + 402 y = 3.5936x + 18859 26588 y y = = -1.9061x 200.9x + +16724 630.36 y = 148.61x + 24597 y y= =54.741x 1.214x + + 6060.1 y = 56.813x +71.316 6466.6 y y = 48.671x + 26822 y= = 1.2573x 53.002x + + 396.92 21849 y y = = 2.6181x 135.59x + + 465.51 y = 132.49x + 27136 23501

0.5118 0.7234 0.522 0.1787 0.6263 0.345

0.5708 0.643 y -0.01x2 ++ 10.25x + 145.2 0.4555 y= = 13190e0.0148x 0.2209x2 244.58x + 18895 y y = = 0.0008x2 0.5341x2 + + 4.6632x 132.55x + 23404 y = 0.7105x2 + 81.877x + + 379.78 28040 0.5636 0.4117 0.3316 0.6901 0.4988 0.4672 0.5219 0.6488 0.4961 0.6257 0.519 0.3897 y 0.0053x2 +3.8609x 548.66x+ +280.43 11147 y= = -2.4512x2 0.2231x2 + + 85.131x + 22370 y 0.0009x2 2.5106x 116.49 y= = 0.0028x2 0.1015x2 + + 105.43x 48.669x + + 11728 19936 y 0.0004x2 +2.9663x 145.51x+ +98.156 10296 y= = -0.0579x2 0.0457x2 + + 81.901x + 19619 y 0.0096x2 + 5.9919x + 106.28 y= = -0.873x2 1.1781x2 + + 397.29x 28.881x + + 7182.4 22770 y -0.5447x2 +1993.2x 60.252x +9502.6 24.594 y= = 17.404x2 54.16x2 - + 112.56x ++ 25847 y 0.0707x2 -433.95x 8.6972x ++ 34831 581.78 y= = 3.642x2 0.3182x2 22.323x + 27074

0.7317 0.4604 0.3639 0.1887 0.5772 0.1684 0.5756 0.5087 0.1777 0.9244 0.6088 0.277 0.6506 0.9258 0.263 0.8034 0.6049 0.2753 0.6181 0.5685 0.2519 0.1287 0.069 0.1992 0.0337 0.1575 0.2278 0.2429 0.3318 0.1788 0.0507 0.032 0.0554 0.0374 0.035 0.0245 0.0466 0.4501 0.0118 0.6261 0.3171 0.1567 0.7896 0.8331 0.3106 0.3996 0.2436 0.0769 0.1268 0.0446 0.0649 0.8867 0.1058 0.285 0.1556 0.4509 0.113 0.1247 0.0754 0.0977

0.4031 0.4956 yy == 168.49Ln(x) 8844Ln(x) -+ -2486.9 62.785 0.4097 y = 7309.3Ln(x) 1342.2 0.9223 0.8063 y y = = 341.79Ln(x) 12699Ln(x) 1043.1 31344 0.7082 y = 13453Ln(x) - 34992 0.9235 0.7788 y = 11366Ln(x) 22765 0.5697 y y= =290.56Ln(x) 14312Ln(x) - 761.09 36333 0.7477 0.8014 0.5281 0.5603 0.6123 0.4021 0.2408 0.2757 0.7167 0.8173 0.7129 0.6383 0.4337 0.423 0.6339 0.0658 0.0011 0.2107 y y = 10730Ln(x) 13656 y= =291.86Ln(x) 11332Ln(x) - 557.17 12849 y y = = 413.35Ln(x) 29097Ln(x) 342.04 y = 21388Ln(x) - 29541 16679

0.0068 0.2514 0.1509 0.185 0.45 y = 180.82Ln(x) 2980Ln(x) + + 375.84 0.2319 yy == 3534.9Ln(x) +21275 25116 0.6259 0.2544 y y = = 308.34Ln(x) 11192Ln(x) 1540.2 46587 0.4015 y = 9915.9Ln(x) - 39193 0.7837 0.8304 0.6071 0.3285 0.2241 0.0732 0.7304 0.116 0.2273 0.8818 0.8315 0.7057 y y= =285.59Ln(x) 14980Ln(x) -- 1262.4 68310 y = 9121.3Ln(x) 27860 y y = = 163.98Ln(x) 5065.5Ln(x) 390.97 3688 y = 7005.4Ln(x) - 12806 y y= = =5971.7Ln(x) -148.79Ln(x) 18871Ln(x) 38275 1019 y ++ 12197 y y = 16439Ln(x) 66989 y= =488.15Ln(x) 19224Ln(x) - 2244.6 82641

y 0.2337x2 3.8478x+ 820.81x +28893 543.61 33495 y= = 41.935x2 6.123x2 +-26.921x yy y== =149.26Ln(x) 7644Ln(x) ++ 118.38 0.3318 0.173 y 0.0653x2 1.3898x 1957.6 + 11277Ln(x) -1957.6 12590 0.4653 y= = 7644Ln(x) 1.0147x2 -+ 16.115x + 513.14 26631 y y = = 52.951Ln(x) 1689.6Ln(x) + + 263.92 17381 0.1958 0.0317 y = 0.1061x2 2.9767x2 9.2704x 490.18x + 584.61 y = 4772.8Ln(x) + 16815 0.159 y = 1.2221x2 - 58.855x + 35759 31283 y y= =-34.956Ln(x) 3241.7Ln(x) + + 21706 22148 516.78 0.1369 0.0331 y 0.0995 y= = -0.2605x2 0.9751x2 -+ 69.854x2 28.536x 1671.5x ++ 597.12 35411 y = 4652.2Ln(x) 237.05x 30410 0.0723 0.4359 0.0601 0.5126 0.3186 0.3415 0.5991 0.574 0.3851 0.3409 0.1622 0.21 0.6291 0.2828 0.1305 0.7602 0.4549 0.5341 0.2583 0.4162 0.2544 0.4634 0.4257 0.3595 y -0.8791x2 + 2052.7x 96.539x + + 36193 244.5 y= = 81.635x2 1.1084x2 + 55.909x + 32483 y -2E-06x2 +0.2938x 30.161x +199.94 8471 y= = -0.0074x2 0.0009x2 + + 4.458x ++ 22194 y -5E-07x2 14.537x 10410 y= = 0.0018x2 0.0038x2 + + 0.4061x 6.7949x + + 167.82 21290 y = -9E-05x2 -0.0073x2 + + 0.5107x 26.054x + + 322.42 y = -0.0011x2 + 10.653x + 16239 25092 y 0.0265x2 6.4468x 626.23 y= = 1.1263x2 0.5319x2 +203.72x 3.2542x + +34160 30959 y = -0.0001x2 0.0836x2 + 56.676x 1.3869x + + 34781 35.212 y = 0.0254x2 + 12.78x + 19956 y -0.0015x2 +38.802x 2.1515x+ 94.407x +25576 303.47 22512 y= = -0.0765x2 0.0071x2 + y -0.0025x2 +40.802x 3.0113x+ 176.65x +30168 480.26 27473 y= = -0.2166x2 0.1822x2 +

0.4102 0.4882 0.3099 yy = = 76.713Ln(x) 4733Ln(x) + + 9053.7 132.18 0.1526 y = 6586.7Ln(x) + 2239 0.2408 0.3022 y y = = 151.32Ln(x) 9273.6Ln(x) + + 246.24 10556 0.4286 y = 9246Ln(x) + 9496.9 0.5065 0.3352

0.2766 0.501 y y= =7209.8Ln(x) 179.9Ln(x) -+ 1321.3 0.2785 y = 9804.2Ln(x) -234.25 16214 0.4909 0.5447 y y = = 124.77Ln(x) 5864.2Ln(x) + + 213.75 16222 0.5966 y = 7760.4Ln(x) + 4167.9

76

S. S. NO. NO.
1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20

Predicting Total Payment Based on Engineer Hrs. Predicting Based on Engineer Hrs. Predicting Total Total Hrs Payment Based on Tech III Hrs. Predicting Based on Tech III Hrs. Predicting Total Total Hrs Payment Based on Tech II Hrs. Predicting Total Hrs Based on Tech II Hrs. Predicting Total Payment Based on Tech I Hrs.

PREDICTING TOTAL PAYMENT WITHOUT OUTLIERS MISCELLANEOUS PREDICTING TOTAL MAH HOURS WITHOUT OUTLIERS MISCELLANEOUS TITLE LEANER LOGARITHMIC TITLE LEANER LOGARITHMIC 2 EQUATION R EQUATION R2 2 EQUATION R EQUATION R2
y = 338.01x + 11936 y y= = 8.5322x 447.35x + + 172.43 8925.8 y = 14.804x + 179.48 y = 190.2x + 8047.5 yy==4.7654x 123.85x++216.29 5394 y= = 123.65x 2.9786x + + 6830.4 80.893 y y= = 222.82x 2.9458x + + 7455.8 122.59 y y= = 1826.2x 8.7197x + + 807.51 87.102 y y= = 155.17x 42.847x + + 8178.4 69.761 y y = 3.3985x + y = 697.25x + 215.42 11699 y y= = 11.612x 438.55x + + 394.94 3709.4 y = 11.279x + y = 241.28x + 107.09 15698 y y= = 6.395x 483.7x + + 338.79 13487 yy == 24.062x + 297.98 2214x + 20253 y= =10.706x 48.944x+ +9082.9 499.2 y y y= = 0.3367x 26.602x + + 188.23 8434.8 y y= = 0.5787x 27.906x + + 164.54 13006 yy==0.5994x + 262.47 193.9x + 13081 y +8802.1 327.29 y= =4.8309x 43.32x + y y= = 1.1132x 53.514x + + 105.81 10854 y y= = 1.7212x 105.05x + + 213.93 16391 y = 1.535x + 403.45 0.7975 0.8144 0.6392 y = 9248.1Ln(x) - 5696.6 y = 258.33Ln(x) - 217.63

Predicting Total Total Payment Hrs Based on Tech I Hrs. Predicting Based on AutoCAD Hrs. Predicting Total Total Payment Hrs Based on AutoCAD Hrs. Predicting Based on Plotter Hrs. Predicting Total Total Payment Hrs Based on Plotter Hrs. Hrs. Predicting Based on Clerical Predicting Total Total Payment Hrs Based on Clerical Hrs. Predicting Based on Total Survey Hrs. Predicting Based on Total Survey Hrs. Survey Hrs. Predicting Total Total Hrs Payment Based on 3 Man Crew Predicting Based on 3 Man Survey Hrs. Hrs. Predicting Total Total Hrs Payment Based on 2 Crew Man Crew Survey Predicting Based on 2 Man Crew Survey Hrs. Hrs. Predicting Total Total Hrs Payment Based on Registered Surveyor Predicting Based on Registered Surveyor Hrs. Predicting Total Total Hrs Payment Based on Survey Technician Hrs. Predicting Based on Survey Technician Predicting Total Total Hrs Payment Based on Survey Office Hrs. Staff Hrs. Predicting Total Hrs Based on Survey Office Staff Hrs. Predicting Total Payment Based on Mileage Cost. Predicting Based on Mileage Cost.Cost. Predicting Total Total Hrs Payment Based on Printing Predicting Based on Printing Cost. Predicting Total Total Hrs Payment Based on Expenses. Predicting Based on Expenses. Predicting Total Total Hrs Payment Based on Total Prelim Hrs. Predicting Total Hrs Based on Total Prelim Hrs. Hrs. Predicting Total Payment Based on Total Design Predicting Based on Total Design Hrs. Predicting Total Total Hrs Payment Based on Total Construction Hrs. Predicting Total Hrs Based on Total Construction Hrs. Predicting Total Payment Based on Total As Built Hrs. Predicting Total Hrs Based on Total As Built Hrs.

0.6291 0.5955 y = 4883Ln(x) + 2798.3 0.7411 33.791 0.3112 3.2784x + +29724 258.02 0.8511 yy==156.48Ln(x) 10693Ln(x) + - 30453 0.7407 y y= = 0.0046x2 0.5439x2 + - 154.63x 0.9516 y 0.8402 y y= -0.0023x2 + + 180.77x 4.0807x + + 3008.4 1.8518 0.8121 y= =383.57Ln(x) 10084Ln(x) - 1254.5 26121 0.7246 = -0.2524x2 0.876 y 0.8045 y y= = 3.3005x2 -0.0021x2-+ 3.8242x 71.716 0.8279 y= = 350.15Ln(x) 9057.5Ln(x) -- 1055.4 12481 0.7669 477.67x ++ 34526 0.901 y y= = -43.139x2 -0.0201x2 + + 3304.1x 11.673x +8224.3 26.062 0.7113 y= =272.05Ln(x) 24265Ln(x) - 408.76 33909 0.6982 0.7303 y 0.619 y y= = 4386.7Ln(x) 521.68Ln(x) + - 628.51 0.5394 y = 1.1555x2 -0.329x2 + 55.613x + - 18.875 0.6219 3536.9 0.2439 y= - 202.13x 27128 0.6148 y = 89.44Ln(x) + 136.03 0.215 y = 0.016x2 1.4862x + 459.92 0.6955 y = 8144.6Ln(x) + 6164.4 0.4743 y = 10.336x2 - 335.09x + 24760 0.3307 283.74 0.2567 8.1097x + +29090 455.34 0.6867 yy==147.49Ln(x) 14821Ln(x) + - 29472 0.6234 y y= = 0.0294x2 7.0248x2 + - 520.29x 0.6512 0.2464 y y= = 416.96Ln(x) 5095.8Ln(x) +880.33 10547 0.3002 y = 110.24Ln(x) + 254.95 0.5843 y = 6033.8Ln(x) + 8459.1 0.5601 == 265.33Ln(x) ++ 121.94 0.3978 yy 6223.3Ln(x) 8972.6 0.1487 0.6414 0.631 0.8477 y y= = 180.3Ln(x) 10044Ln(x)+- 525.13 45038 y = 258.92Ln(x) 1171.8 y = 6918.1Ln(x) - 19234 0.6375 0.3038 0.3179 0.4812 0.4053 0.32 0.1671 0.5353 0.561 0.5128 0.6701 0.5923 0.6642 0.2378 0.2542 0.5326 0.7177 0.5972 0.5084 0.6644 y 1.4152x + +26765 415.6 y= = 0.0622x2 3.1066x2 + - 164.91x y 483.2 y= = 0.1689x2 12.165x2 - 6.7375x 462.4x ++ 26034 y 12.152x + +26170 471.29 y= = 0.1456x2 5.0481x2 + - 108.95x y 59.491x ++ 704.05 y= = 8.4705x2 -0.0016x2-+ 17.247x 5820.1 y y= = -7E-05x2 -0.0025x2++0.5562x 23.436x++88.534 7654.9 y = -2E-05x2 + 0.6099x + 157.37 y = -0.0083x2 + 33.127x + 6456.2 y 1.1696x 124.33 y= = -0.0003x2 3.8303x2 -+ 385.11x ++ 28655 y y= = 0.0591x2 0.1559x2 - 4.4312x 110.23x + + 599.79 34103 y 1.4987x 49.694 y= = -0.0004x2 0.0861x2 -+ 45.869x ++ 28518 y y= = 0.0028x2 3.6496x2 - 0.4001x 928.89x + + 494.65 40985

0.4924 y = 9396.4e0.0139x 0.5231 y =y224.79e0.0128x = -1.548x2 + 495.4x + 10168 y = -0.0019x2 + 13.412x + 236.95 0.3559 y = 1.7161x2 - 238.21x + 27519

POLYNOMIAL POLYNOMIAL EQUATION EQUATION

0.5526 0.6254 0.4502 0.5338 0.0752 0.7605 0.0378 0.9643 0.7366 0.8829 0.0741 0.9165 0.7396 0.6358 0.0896 0.265 0.0678 0.1578 0.063 0.2698 0.0113 0.1024 0.0223 0.3611 0.0096 0.0898 0.665 0.6684 0.6197 0.8582 0.5943 0.7562 0.0327 0.0644 0.1646 0.8138 0.0724 0.1788 0.0218 0.014

R2 R2

0.8579 0.7226 y y= = 213.95Ln(x) 7313.3Ln(x) -- 692.63 23794 0.6789 y = 298.51Ln(x) 1138.2 0.3703 y = 6494.5Ln(x) + 1094.1 0.3723 9.8834 0.754 yy==166.84Ln(x) 14346Ln(x) + - 57997 0.801 0.7602 0.7067 0.8136 yy == 298.17Ln(x) 1164.1 9891Ln(x) --25368

y= =8288.7Ln(x) 216.08Ln(x)--5556.1 513.4 y 0.0102 y = 92.354Ln(x) + 152.06 0.0458 y = -0.1429x2 + 5.6231x + 529.97

77

PREDICTING TOTAL PAYMENT WITHOUT OUTLIERS SUBDIVISION PREDICTING TOTAL MAN HOURS WITHOUT OUTLIERS SUBDIVISION S. NO.
1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20

TITLE TITLE

LEANER LEANER EQUATION EQUATION

2 R R2

LOGARITHMIC LOGARITHMIC EQUATION EQUATION

y= =230.24Ln(x) 13724Ln(x) - 162.07 16550 y y y= = 7705.3Ln(x) 168.01Ln(x) - 521.53 50.138 y = 12980Ln(x) 30720 y = 326.96Ln(x) - 911.33 y= =396.89Ln(x) 16083Ln(x) - 1194.9 42904 y y= = 269.52Ln(x) 10801Ln(x) y - 7347.7 326.76
2 R R2

POLYNOMIAL POLYNOMIAL EQUATION EQUATION

2 R R2

Predicting Based on Engineer Predicting Total Total Payment Hrs Based on Engineer Hrs. Hrs. Predicting Total Payment Based on Tech III Hrs. Predicting Total Hrs Based on Tech III Hrs. Predicting Based on Tech II Hrs. Predicting Total Total Payment Hrs Based on Tech II Hrs. Predicting without MEP Based on Tech I Hrs. Predicting Total Total Payment Hrs Based on Tech I Hrs. Predicting Based without Hrs. MEP on Autocad Hrs. Predicting Total Total Payment Hrs Based on AutoCAD Predicting Based on Plotter Predicting Total Total Payment Hrs Based on Plotter Hrs. Hrs. Predicting Based on Clerical Predicting Total Total Payment Hrs Based on Clerical Hrs. Hrs.
Predicting Total Payment without MEP Based on Total Survey Hrs.

y= = 4.9131x 282.26x + + 410.39 18385 y y = 231.49x + y = 3.0014x + 26749 589.8 y= = 3.6353x 155.19x + + 310.77 16980 y y 10140 y= = 143.11x 3.216x ++125.26 y 11025 y= = 164.85x 3.337x ++199.42 y= = 6.2839x 276.81x + + 300.98 15983 y yy==1648.6x + 216.4 14770 40.47x + y= = 2.3731x 148.65x + + 343.59 12122 y y 16882 y= = 733.11x 1.831x ++528.61 y= = 5.2266x 238.76x + + 409.63 14513 y y= = 3.7948x 155.32x + + 545.98 25823 y y= = 13.143x 213.79x + + 607.03 31532 y y= = 19.941x 736.71x + + 696.02 32985 y y= = 0.1797x 9.808x ++16610 y 387.9 y= = 0.5033x 23.849x + + 278.08 13133 y y= = 0.2495x 1.3356x + + 530.37 32050 y y 28776 y= = 117.07x 2.689x ++567.53 y= = 1.1268x 53.419x + + 199.49 11610 y y= = 1.7184x 56.647x + + 371.05 22711 y y= = 3.0153x 136.95x + + 497.25 22890 y

0.46 0.5742 0.1571 0.0746 0.3616 0.5085 0.7404 0.9068 0.7166 0.8528 0.5111 0.7781 0.2464 0.4453 0.8222 0.636 0.6071 0.0033 0.6254 0.637

0.4113 y = 18313e0.0083x 0.4478 y = -0.0022x2 + 5.4375x + 386.66 y = 2.9807x2 - 212.85x y = -0.0214x2 + 3.9445x + + 35351 627.75 0.281 y-0.0063x2 = 0.298x2 + +5.2646x 40.388x + +232 25829 0.381 y = 0.4041 y -0.0003x2 = 0.0675x2 66.325x 20459 0.5754 y = ++ 3.3942x ++ 105.61 0.4559 y -0.0012x2 = 0.1178x2 32.209x 28550 0.6598 y = ++ 3.9322x ++ 147.94

0.3768 0.5793 0.0754 0.0286 0.1834 0.5334 0.3494 0.9094 0.1685 0.8623 0.2354 0.7874 0.082 0.4661 0.4436 0.5741 0.4695 0.5602 0.4271 0.6255 0.1442 0.3376 0.0556 0.1664 0.0234 0.2189 0.0944 0.4564 0.1945 0.7177 0.0198 0.3161 0.0584 0.0851 0.1939 0.8373 0.1665 0.5064 0.2005 0.2507

Predicting Total Hrs Based on Total Survey Hrs. Predicting Based on 3 Crew Man Crew Survey Predicting Total Total Payment Hrs Based on 3 Man Survey Hrs. Hrs. Predicting Based on 2 Crew Man Crew Survey Predicting Total Total Payment Hrs Based on 2 Man Survey Hrs. Hrs. Predicting Based on Rigestered Surveyor Predicting Total Total Payment Hrs Based on Registered Surveyor Hrs. Hrs. Predicting Based on Survey Technician Predicting Total Total Payment Hrs Based on Survey Technician Hrs. Hrs. Predicting Total Payment Based on Survey Office Staff Hrs. Predicting Total Hrs Based on Survey Office Staff Hrs. Predicting Based on Mileage Predicting Total Total Payment Hrs Based on Mileage Cost. Cost. Predicting Based on Printing Predicting Total Total Payment Hrs Based on Printing Cost. Cost. Predicting Based on Expenses. Predicting Total Total Payment Hrs Based on Expenses. Predicting Based on Total Prelim Predicting Total Total Payment Hrs Based on Total Prelim Hrs. Hrs.
Predicting Total Payment without MEP Based on Total Design Hrs.

0.3317 y -0.0072x2 = 1.7837x2 117.21x++246.8 32218 0.5855 y = + -7.7772x y= = 362.63Ln(x) 16759Ln(x) y 0.3722x2 = 88.361x2 - 1460.4x 38535 y - 5071.6 156.55 0.2385 0.3516 y = + 31.472x ++ 255.88 y y 0.0029x2 = 0.2553x2 + 2.8787x + 25317 y= = 6735.2Ln(x) 134.69Ln(x) + + 3601.4 124.09 0.2774 0.3201 y = + 1.0505x + 459.29 y= = 223.54Ln(x) 8444.1Ln(x) + + 161.37 13161 0.1993 y 0.0717x2 = 6.2052x2 + 16.648x + 29019 y 0.3754 y = + 4.6881x + 569.59

y - 207.8 18443 0.4015 y 0.0063x2 = 0.8005x2 + 29.455x + 24976 y= =12710Ln(x) 257.1Ln(x) 0.3338 y = + 3.5191x + 479 0.2252 y= = 126.27Ln(x) 4682.7Ln(x) + + 343.09 18787 0.0126 y 0.0007x2 = 0.9598x2 - 17.333x 30908 0.5081 y 0.274 y = + 3.5619x ++ 560.1 0.1049 y y 3248.8Ln(x) 28021 0.1035 0.0551 y y -1.0606x2 + 309.81x + 31050 0.2112 == 131.52Ln(x) ++ 513.65 == 0.1923x2 + 3.1556x + 650.07 0.0027 0.2189 0.3855 0.4084 yy==208.67Ln(x) + 649.3 38246 123.4Ln(x) + y y= = 10898Ln(x) 215.07Ln(x)--45102 851.7 0.0002 y 0.0215x2 = 0.1928x2 + 128.16x + 34286 0.1854 y = + 19.252x + 698.4 0.3451 y -4E-05x2 = 0.0024x2 - 3.5943x 31888 0.4172 y = + 0.3583x ++ 260.16 0.2912 y -3E-05x2 = 0.0101x2 - 5.9581x 30186 0.4221 y = + 0.5872x ++ 235.63

Predicting Total Hrs Based on Total Design Hrs. Predicting Based on Total Construction Predicting Total Total Payment Hrs Based on Total Construction Hrs. Hrs. Predicting Based on Total AsBuilt Predicting Total Total Payment Hrs Based on Total As Built Hrs. Hrs.

0.5713 7730Ln(x) --16881 0.7002 y y == 178.84Ln(x) 450.67 0.0118 y = 5171.3Ln(x) + -0.0009x2 + 6.1624x + 32206 0.2994 y = 189.67Ln(x) - 555.11 445.3 0.1086 0.2595 y y == 3E-05x2 + 0.1532x + 564.72 0.1093 y= = 130.43Ln(x) 5436.9Ln(x) + + 254.86 15984 0.1143 y 0.001x2 = 0.1643x2 + 67.921x + 31675 0.1321 y 0.141 y = + 1.9247x + 635.88 0.714 y - 2351.6 104620 0.5557 y-0.0002x2 = 0.026x2 + +1.4693x 3.1742x + +106.15 26779 0.8152 y= =22993Ln(x) 512.6Ln(x) 0.765 y = 0.2397 0.5726 0.6359 0.5776 y= =314.43Ln(x) 11860Ln(x) - 878.22 26199 y y= = 271.35Ln(x) 10945Ln(x) y - 9853.8 348.15 0.2258 y 0.001x2 = 0.0105x2 + 42.524x + 25833 0.4364 y = + 1.0648x + 445.14 0.4268 y 0.0002x2 = 0.2421x2 + 9.2227x + 32161 0.5138 y = + 2.3638x + 600.86

78