Comparison of 2018 & 2021 Survey Data from Bailey Library Clients Elijah McClymonds, Jana Asher, & Amanda Goodrick Department of Mathematics & Statistics, Slippery Rock University Introduction: Every three years, the SRU Bailey Library participates in a national survey that measures library user satisfaction, library quality, and library functionality. This survey was made available by the Bailey Library staff for the 2021 winter session DataCamp run through the SRU Mathematics/Statistics Department. This constructed study uses the survey data with over 1,000 respondents from the years 2018 and 2021. Each yearly dataset measures the minimum level(mn), desired level(de), and perceived level(pr) of a corresponding question. This data is further made measurable by the two included variables Adequacy Gap(ad) and c Superiority Gap(su). These two variables are calculated using the corresponding user’s perceived level minus the minimum level and the corresponding user’s perceived level minus the desired level, respectively. The data is also divided into three distinct themes which are Affect of Service(AS), Information Control(IC), and Library as a Place(LP). Affect of Service questions deal with staff performance, Information Control questions deal with access to library resources, and Library as a Place questions deal with the environment and atmosphere the library creates for a user. Note that there are separate questions that are unique to a year’s survey (i.e. L17359 is unique to the 2021 survey). There are many descriptive variables that reflect the user such as major, age, and position as well. Objectives: For this study, we want to: i. ii. compare 2018 and 2021 survey results determine the most meaningful cquestions in the most recent (2021) survey iii. Investigate survey responses based on major (2021) Results: 2021 Cluster Analysis 2018 Vs 2021 Analysis ✓K-means clustering is a technique in which we place each observation in a dataset into one of K clusters (in this case k=5). ✓This K-means cluster is run comparing the means across different majors at SRU. ✓Cluster 1 contains Engineering, Other, and Physical and Health majors ✓Cluster 2 contains the Homeland/Corporate Security major ✓Cluster 3 contains Accounting/Economy/Finance, Art/Dance/Theater, Biology, Criminology, Elementary Ed/Early Childhood, Exercise & Rehabilitative Sci, Healthcare Adm & Info Systems, Management & Marketing, Mathematics/Statistics, Parks, Conserve, & Rec Therapy, Physical Therapy, Physician Assistant, Psychology, Safety ✓ Cluster 4 contains the Music, ✓Cluster 5 contains the Computer Science, Management, Special Occupational Therapy, and English, Geography/Geology/Environment, Education, and Strategic Comm Public Health/Social Work History, Secondary Ed/Foundations Ed and & Media majors. majors Undecided/Explored Studies majors Mean (2018) Variable Label Question 8 - The printed library materials I 0.0486 IC07_su need for my work (superiority gap) Question 11 - The electronic information 1.0672 IC10_ad resources I need (adequacy gap) Question 11 - The electronic information -0.1491 IC10_su resources I need (superiority gap) Question 17 - Modern equipment that lets me easily access needed information (adequacy 0.7177 IC14_ad gap) Question 17 - Modern equipment that lets me easily access needed information (superiority -0.2744 IC14_su gap) Question 22 - Willingness to help users 0.9382 AS18_ad (adequacy gap) Average desired expectation for first 27 7.6079 AvgDes questions Average perceived experience for first 27 7.349 AvgPer questions Average perceived experience regarding the 7.3342 D2AvgPer library's information control Average adequacy gap regarding the library's 0.8472 D2AdqGap information control Average superiority gap regarding the library's -0.2722 D2SupGap information control Mean (2021) P-Value -0.3415 0.0024 0.8479 0.0017 -0.3614 0.0007 0.231 0.0001 -0.941 0.0001 0.6307 0.0083 7.4434 0.0019 7.1539 0.0003 7.0959 0.0001 0.6859 0.0031 -0.4338 0.0007 From the table constructed from the two-sample t-test, we can note: ✓ Multiple variable means dropped significantly in value from 2018 to 2021. ✓ The survey data indicates overall perception and expectation levels went down from 2018 to 2021 and the amount of negligence people are willing to put up with regarding the library’s information fell. ✓Specific questions indicate the areas the Bailey Library should improve upon. 2021 Question Analysis Survey Data Deficiencies Methodology: For i: We will use SAS software to perform two sample t-tests. A two-sample t-test is a statistical hypothesis test that compares a mean of a sample across two classes to see if they yield different results. For every variable, we will compare the measurable variables’ means across two different years (2018 vs 2021). Variable IC02_ad IC07_ad IC14_ad AS22_ad L17363_ad IC05_su IC10_su IC14_su L17363_su Label Question 2 - Making electronic resources accessible from my home or office (Adequacy) Question 8 - The printed library materials I need for my work (Adequacy) Question 17 - Modern equipment that lets me easily access needed information (Adequacy) Question 27 - Dependability in handling users' service problems (Adequacy) Question 26 - Access to innovative technology in Bailey Library (Adequacy) Question 5 - A library Website enabling me to locate information on my own (Superiority) Question 11 - The electronic information resources I need (Superiority) Question 17 - Modern equipment that lets me easily access needed information (Superiority) Question 26 - Access to innovative technology in Bailey Library (Superiority) Survey Data Proficiencies Mean Difference P-value 0.1389 0.011 0.1602 0.0113 0.5727 0.0001 0.2258 0.0001 0.3181 0.0001 0.1353 0.0055 0.074 0.0176* 0.5937 0.0001 0.3652 0.0001 Variable Label LP03_ad AS04_ad Question 3 - Library space that inspires study and learning (Adequacy) Question 4 - Giving users individual attention (Adequacy) Question 14 - Instruction sessions conducted by librarians helping me to better understand the research process (Adequacy) Question 15 - A comfortable and inviting location (Adequacy) Question 16 - Employees who deal with users in a caring fashion (Adequacy) Question 25 - Community space for group learning and group study (Adequacy) Question 13 - Employees who have the knowledge to answer user questions (Superiority) Question 16 - Employees who deal with users in a caring fashion (Superiority) L17361_ad LP12_ad AS13_ad LP21_ad AS11_su AS13_su Mean Difference P-Value 0.1718 0.0001 0.1177 0.0184* 0.2401 0.0043 0.14 0.0861 0.1474 0.1111 0.117 0.005 0.0032 0.0081 0.0093 0.0001 ✓ The table constructed on the left indicates the questions that weigh the overall average down and by extension need improvement. The table on the right indicates the questions that weigh the overall average up and by extension deserve recognition. ✓ Information Control yields the most concern for the survey respondents, while employee service and the library environment yield the best results for the survey respondents. For ii: We will continue using SAS software to perform paired t-tests. A paired t-test is another statistical hypothesis test that generally compares a mean of one variable to a second variable. c In this circumstance, we will compute the mean of each measurable variable and compare the value to the mean of the corresponding hierarchal variable. In other words, ad will match up with AdqGap and su will match up with SupGap. Conclusion/Outcomes: Future Work: For iii: ✓ The average respondent survey data yielded statistically significant results only for 2021 mean data less than 2018 mean data. Results suggest that desired expectations along with minimum expectations have decreased over time. We will use the statistical program R to perform a k-means data cluster. This test indicates if students answer similarly based on major. We plot the data clusters to visually illustrate the classification of students’ responses by major. ✓ Library user survey data could alternatively be analyzed using a time series method. If we have access to more of the data from the past, we could try and analyze the data to determine whether there is a trend. ✓ The access to library resources and information is a growing problem as documented in the Data Deficiencies table. ✓ Employee interaction and the environment at the Bailey library are both strong areas for the library. ✓ The cluster analysis indicated that student responses across several majors were similar (See Cluster 3). ✓ The results of the paired t-test can be extended further to see the driving forces regarding the thematic labels of each question. ✓ The data clustering methods used could be implemented and optimized greater than the generic k-means cluster.