You are on page 1of 27

SDSU

Aztec Recreation Center


Membership Survey
This report contains the statistical analysis of a membership survey conducted by a team of SDSU students.

A r i e l a E n g b e r C a r l e n n S m i t h Q u e n t i n A n d e r s o n T y l e r B o o t h M i k e J o n e s

Key Centrality and Dispersion


Statistics for six different groups of people on their average importance ratings for cleanliness: Males with a membership less than 1 year, 1 to 3 years, and greater than 3 years and females with a membership less than 1 year, 1 to 3 years and greater than 3 years. The groups are divided both by length of membership and gender. Membership Length (years) 1 = Males < 1 2 = Males 1 - 3 3 = Males > 3 4 = Females < 1 5 = Females 1 - 3 6 = Females > 3

INTERPRETATION The results of the box and whisker plot suggest that women rate the importance of cleanliness higher than men in all levels of membership length. This conclusion can be made based on the small range of womens rates and the higher medians in all levels of membership length, demonstrating that the overall expectation for gym cleanliness is higher for women. The range of ratings on cleanliness was found to be much larger for men, especially in the case of men with memberships from 1 to 3 years and 3+ years. This may be attributed to the

fact that males who have been at the ARC less than a year may be SDSU freshman with free memberships and do not renew their membership at the ARC when the free trial ends. Members who return after their free freshman membership expires (1-3yrs, >3yrs) are less concerned with cleanliness of the ARC.

Categorized Groups
A single chart categorizing people into two groups - Satisfied and Unsatisfied with further categorization by gender. The category unsatisfied contains survey respondents who indicated satisfaction of 1, 2, or 3, and the category satisfied is composed of respondents who answered with a 4 or 5.

INTERPRETATION

The results of the stacked bar chart suggest that approximately 55% of the unsatisfied individuals were women, with approximately 45% of unsatisfied individuals being men. The chart also suggests that out of the total satisfied individuals, approximately 55% were men and approximately 45% of satisfied individuals were women. These results may be attributed to an importance of cleanliness to women being higher overall than men, in conjunction with higher expectations for cleanliness in a gym than men.

Descriptive Information
GENDER
Results: Gender Male Valid Female Total Missing System Total Frequency 188 210 398 72 470 Percent 40.0 44.7 84.7 15.3 100.0 Valid Percent 47.2 52.8 100.0 Cumulative Percent 47.2 100.0

The above frequency table explains that out of a total of 398 valid respondents, there were 188 males and 210 females.

The bar chart shows gender breakdown based on the percentages of valid participants. 47.2% of respondents were male and 52.8% were female. This may be a result of the gender ratio of the SDSU population, as there is a higher proportion of female students than male students.

PRIMARY TIME WORKING OUT


Percent Valid Percent Cumulative Percent 6am-11am 18.5 21.2 21.2 11am-4pm 22.8 26.1 47.3 Valid 4pm-12am 43.2 49.5 96.8 12am-6am 2.8 3.2 100.0 Total 87.2 100.0 Missing System 12.8 Total 100.0 Out of a total of 410 valid respondents, 87 preferred to work out from 6am to 11am, 107 preferred to work out from 11am to 4pm, 203 preferred to work out from 4pm to 12 am, and 13 preferred to work out from 12am to 6am. .

Frequency 87 107 203 13 410 60 470

The bar chart shows gender breakdown based on the percentage of valid participants. The majority (49.5%) of people prefer to work out from 4pm to 12am. This may be explained by the high student membership base at the ARC, because this may be the time students are attending class, making it convenient for them to also go to the gym while they are on campus.

MEMBERSHIP TYPE
SDSU Student Valid Affiliate/Faculty/Staff/Alumni Community Total Missing Total System Frequency 286 89 24 399 71 470 Percent 60.9 18.9 5.1 84.9 15.1 100.0 Valid Percent 71.7 22.3 6.0 100.0 Cumulative Percent 71.7 94.0 100.0

The frequency table explains that out of a total of 399 valid respondents, there were 286 SDSU student memberships, 89 Affiliate/Faculty/Staff/Alumni memberships, and 24 Community memberships.

The bar chart shows the breakdown of membership type based on the percentages of valid participants. The largest percentage is SDSU student memberships at 60.9%, while the smallest percentage is Community members at 6% of all memberships. This is most likely the result of the location and marketing of the ARC, as it is located on campus (close to where many students live) and is one of the features that the school promotes to incoming freshman and students.

Sample Information
LEAST IMPORTANT FEATURE

INTERPRETATION Staff greeting upon entrance was rated the least important feature of the ARC, as the average respondent rating was approximately 2.9 out of 5. This might be because of the ARCs large membership base, as members may not expect to be greeted when there are so many members as compared to the staff.

FAVORITE TYPE OF MUSIC

INTERPRETATION

Approximately 21% of valid respondents stated that PUMP!! was their music preference for the gym. This may be because members like to work out to high intensity music.

MOST POPULAR TV STATIONS


1. ESPN 2. Comedy Central 3. MTV ARC Members Top 10 TV Shows (frequency)

INTERPRETATION ESPN, Comedy Central, and MTV were chosen most often by ARC members when asked what TV shows they would like to watch at the gym. ESPN was chosen 193 times, Comedy Central was chosen 150 times and MTV was chosen 132 times. This may be because people who are sports fans also maintain an active lifestyle through having an ARC membership.

Difference Between Groups


QUESTION
Is there a difference in the primary time of day that men and women work out? Similarly, is there a difference in the primary time of day that students and faculty members work out? This research question requires analysis to determine whether there is a relationship between two categorical variables; gender (nominal) and workout time (ordinal). To determine this relationship, we chose to run a chi- square test through crosstabulation at the .05 level to examine the expected results as compared to the actual results. Gender * Primary time working out Crosstabulation Count Male Gender Expected Count % of Total Count Female Expected Count % of Total Count Total Expected Count % of Total 44a 39.5 11.4% 40a 44.5 10.4% 84 84.0 21.8% Primary time working out 6am 11am 11am 4pm 46a 49.8 11.9% 60a 56.2 15.6% 106 106.0 27.5% 4pm 12am 91a 91.7 23.6% 104a 103.3 27.0% 195 195.0 50.6% 181 181.0 47.0% 204 204.0 53.0% 385 385.0 100.0% Total

Each subscript letter denotes a subset of Primary time working out categories whose column proportions do not differ significantly from each other at the .05 level.

The crosstabulation clearly demonstrates that there is no notable difference between the actual data and the expected data, implying a nonmonotonic relationship between gender and workout time. Chi-Square Tests Pearson Chi-Square Likelihood Ratio Linear-by-Linear Association N of Valid Cases Value 1.538a 1.537 .436 385 df 2 2 1 Asymp. Sig. (2- sided) .464 .464 .509

a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 39.49.

The chi-squared difference test portrays the significance of the difference between these two variables. With the p-value (0.464) being much greater than the beta value (.05), we fail to reject the null hypothesis that gender and workout time are independent of one another.

students 1 faculty 2 * Primary time working out Crosstabulation Count 1.00 Expected Count students 1 faculty 2 % of Total Count 2.00 Expected Count % of Total Count Total Expected Count % of Total Primary time working out 6am 11am 11am 4pm 4pm 12am 49a 61.6 13.5% 32a 19.4 8.8% 81 81.0 22.3% 77b 77.6 21.2% 25b 24.4 6.9% 102 102.0 28.0% 151b 137.7 41.5% 30b 43.3 8.2% 181 181.0 49.7% 277 277.0 76.1% 87 87.0 23.9% 364 364.0 100.0% Total

Each subscript letter denotes a subset of Primary time working out categories whose column proportions do not differ significantly from each other at the .05 level. Unlike gender, the crosstabulation taking membership type into account actually does demonstrate that there is a notable difference between the actual data and the expected data, implying a monotonic relationship. Chi-Square Tests Pearson Chi-Square Likelihood Ratio Linear-by-Linear Association N of Valid Cases Value 16.207a 15.489 15.684 364 df 2 2 1 Asymp. Sig. (2- sided) .000 .000 .000

a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 19.36. With the p-value (.000) being less than the beta value (.05), we reject the null hypothesis that membership type and workout time are independent of one another. INTERPRETATION The data indicates that gender does not have an effect on the time of day an individual works out, and that regardless of gender, a higher percentage of people work out later in the day overall. This can be a result of the ARCs high level of student membership and the effect their school schedule plays upon their workout habits. The results also indicate that there is a difference between membership type and workout times. Students tend to work out later in the day, but staff members have no central tendency at any particular time of day. This could possibly be due to students having schedules that typically tend to be in the earlier part of the day (with work and school in consideration) and faculty having more flexible schedules.

QUESTION
On average, do people find the quality of cardio room equipment to be more important than the quality of selectorized (pin loaded) equipment? Is there a difference between their average satisfaction with the performance of cardio equipment and selectorized equipment? This research question requires two analyses to test if ARC members have a different average score on two different continuous variables; cardio room equipment and selectorized (pin loaded) equipment. We chose to run a paired samples t-test with 95% confidence to determine this difference. Paired Samples Test Mean Paired Differences Std. Std. Error Deviation Mean 95% Confidence Interval of the Difference Lower Importance - Quality of cardio room equipment Pair - Importance - 1 Quality of selectorized (pin loaded) weight machines Upper t df Sig. (2- tailed)

.342

.845

.040

.264

.420

8.649

455

.000

The results of the paired samples test shows that there is in fact a difference in the perceived importance in quality of cardio room equipment and selectorized (pin loaded) weight machines. This is clear from the the t- value (8.649) being greater than the z-value (1.96) and the significance (.000) being less than beta (.05). We are 95% confident that there is a statistical difference between the importance in quality of cardio room equipment and that of selectorized (pin loaded) weight machines. The positive range also implies that the perceived importance in quality of cardio room equipment is greater than that of selectorized (pin loaded) weight machines.

Paired Samples Test Mean Paired Differences Std. Std. Error Deviation Mean 95% Confidence Interval of the Difference Lower Performance - Quality of cardio room equipment - Performance - Pair 1 .197 Quality of selectorized (pin loaded) weight machines Upper t df Sig. (2- tailed)

.730

.037

.124

.270

5.301

385

.000

The results of the paired samples t-test shows that, with 95% confidence, that there is in fact a difference in the perceived performance in quality of cardio room equipment and selectorized (pin loaded) weight machines. This is clear from the the t-value (5.301) being greater than the z-value (1.96) and the significance (.000) being less than beta (.05). The positive range of the confidence interval also implies that the perceived performance in quality of cardio room equipment is greater than that of selectorized (pin loaded) weight machines. INTERPRETATION The results of our paired samples t-test shows that on average, people rank the importance and performance in quality of cardio room equipment higher than the importance and performance of quality pin loaded weight machines. This could possibly be a result of a higher proportion of women than men being surveyed, as women reportedly use cardio equipment more than men while men use weight machines more than women, but further analysis would be needed to identify a correlation.

QUESTION
Is there a difference between how SDSU students and Affiliates/Faculty/Staff/Alumni perceive the performance of ARC staff members? Specifically, is there a difference between how these two groups assess staff performance based on: - Staff accessibility - Whether staff greet members upon arrival - Whether staff make members feel welcome - How often members seek staff assistance Because one continuous outcome variable is dependent on a single categorical predictor variable, an independent samples t-test will be used to test if there is a significant difference between the two membership types and their satisfaction with ARC staff members based upon accessibility, whether they greet members upon arrival, whether they make members feel welcome, and with how often members seek staff assistance.

There is only a significant difference between SDSU Students and Faculty with how they perceive performance of ARC staff in regards to being greeted when entering the ARC. We are 90% confident that students will on average feel less satisfied with how they are greeted when they enter the ARC than faculty. Faculty has a higher perception than students do, as signified by the negative confidence interval for that particular variable. It is also signified by the fact that this variables p-value (.001) is less than beta (.10). For the three other variables, we fail to reject Ho and are 90% confident that there is no significant difference between how satisfied students and faculty are with ARC staff member performance. In regards to feeling welcome at the ARC, the p-value (.224) is greater than beta (.10). Satisfaction with being able to find a staff person when needed also shows no significant difference as signified by the p-value (0.839) being greater than beta (.10), as well with how often one wants to find a staff person; p-value of .879 compared to .10.

INTERPRETATION SDSU Affiliates/Faculty/Staff/Alumni have a higher satisfaction level with how they are greeted by ARC staff members than students do. This could be due to the fact that ARC staff members tend to be students, as power roles might influence their likelihood to greet faculty, but further analysis is needed to determine the causality of these results.

QUESTION
Can members monthly hours spent at the gym be predicted by their: - Gender - Favorite T.V. station (ESPN & ESPN 2) - Overall satisfaction - Perceived importance of clean stretching mats - Perceived importance of room temperature To test to see if members monthly hours spent at the gym can be predicted by various other variables, such as gender, favorite TV station, overall satisfaction, perceived importance of clean stretching mats, and perceived importance of room temperature, we will run a linear regression to find a line that best predicts the trends of the data and to test the strength of correlation.

Gender

- - Low R2 value shows weak correlation p-value (.000) compared to beta (.05) implies significance, however positive range for confidence interval actually shows that women will spend between 1.7904.111 hours more than men at the ARC per month (with 95% confidence)

TV Channel

- -

Very low correlation for both TV channels (.004 for ESPN and .001 for ESPN2) Insignificant difference (if any) signified by confidence interval containing 0 in both variables

Overall Satisfaction

- - - Some correlation (.153) There IS a significant difference (p-value .000 < beta .05) Negative confidence interval implies a significant difference as well and that people who are MORE satisfied spend LESS time at the ARC per month (we can predict about 2.679 less hours with 95% confidence)

Importance of Clean Stretching Mats

- Low correlation (.076) - Shows significance (positive confidence interval) - People who place higher importance on clean mats spend more time at the arc per month (by 7.036 hours) INTERPRETATION Though these results may not seem intuitive, they may be explained by those members who are more experienced gym members who more frequently go to the gym, and therefore, they can better assess what they value in a gym than those who do not go as often.

Importance of Room Temperature

- - - INTERPRETATION People who are less satisfied with the temperature of rooms at the ARC may be more frequent gym users, causing them to have a stronger opinion as to how much they value this factor in a gym. Very low correlation (.02) Shows significant difference, p-value (.002) is less than beta (.05) Confidence interval does not contain 0, and has a positive range, implying that people who rank the importance of temperature as very important will spend more time at the ARC per month

Variables and Logic for Inclusion


Variable Description Gender Variable Name Logic for Inclusion

Gender

Through analyzing the demographics of the SDSU population, we have reason to believe that female student concern with body in addition to the high ratio of women to men will predict that women will spend more hours a month at the ARC. Students have generally live closer to the ARC, than faculty who live closer to campus also live closer to the ARC, making it more conven, on average, than faculty, and therefore Members who have held their ARC membership for longer have decided to maintain their membership, implying that they utilize their membership. Therefore, those who have had a longer membership are predicted to spend more time per month than those who have held a shorter membership. Those who are more satisfied with the temperature of the ARC are predicted to spend a longer time at the gym when they do go, and therefore, are more likely to spend more time per month overall than those who are dissatisfied with the temperature of the gym. People who work out in the mornings may have less time to spend while they are at the gym than those who go later in the day, predicting that people who go later in the day will spend more time per month at the gym.

Membership Type

Mem_Type

Membership Length

Mem_Length

Satisfaction with room temperature

Prf_Atm_Temp

Primary time working out

WrkoutTime

Gender


Model Summary Model R R Square
b

Adjusted R Square

Std. Error of the Estimate

.244

.059

.057

5.8805

a. Predictors: (Constant), men 0 women 1 b. Dependent Variable: FAKE DATA. This is a SIMULATED variable. It it the number of hours a respondent spends at the ARC, on average, the previous month.

Low R2 value shows weak correlation p-value (.000) compared to beta (.05) implies significance, however positive range for confidence interval actually shows that women will spend between 1.7904.111 hours more than men at the ARC per month (with 95% confidence) INTERPRETATION - - Women may be predicted to spend slightly more time at the ARC per month than men because they tend to use cardio machines, giving them aerobic workouts over a long amount of time, as opposed to men who more frequently use weight machines for anaerobic workouts that take less time. However, to explore the relationships between gender and workout time further we would need to conduct exploratory research to determine what factors (such as length of typical workout) may be the cause of this weak positive relationship between women and workout time.

Membership Type

- -

Low correlation coefficient (.012) Significant difference (.027<.05) and no 0 in confidence interval. Positive range in confidence interval implies that faculty members spend more time per month at the arc (we can predict about 1.136 more hours with 95% confidence)

INTERPRETATION Staff who choose to be ARC members may spend more time at the gym due to an external factor explained by their choice to become ARC members (such as involvement with sports teams). More research must be done to assess the underlying factors influencing ARC membership to explore the relationship between membership type and monthly hours.

Membership Length

- -

Some correlation (.367) Significance (.000<.05), 0 is not in the confidence interval, positive range in confidence interval implies that individuals who have had a membership for a longer period of time spend more time at the arc per month. Y=.606x + 13.661 Members with longer memberships will spend about 4.452 more hours (with 95% confidence)

Performance and Room Temperature

- - -

Low correlation (.011) Significance (.037<.05) and no 0 in confidence interval Negative range of confidence interval implies that people who are LESS satisfied with the performance of temperature of the rooms spend MORE time at the ARC per month (95% confidence)

INTERPRETATION People who are less satisfied with the temperature of rooms at the ARC may be more frequent gym users, causing them to have a stronger opinion as to how much they value this factor in a gym.

Workout time

- -

Hardly any correlation (.001) Insignificant results, p-value (.500) is greater than beta (.05); also shown through confidence interval range containing 0

You might also like