About the Library: Extended Hours Experiment - 2013

The Thrift Library is open seven days a week for a total of 86 hours.

Hours are determined by usage. The library makes regular half hour counts of the number of people in the library, particularly around opening and closing, and retains them for analysis.

When we detect a trend, we experiment with longer hours, and our users are given the opportunity to "vote with their feet."

Over the years there have been multiple experiments with longer hours which have resulted in the expansion of hours on several occasions, the most notable of which was from 11:00 to 12:00 Sundays through Thursdays.

Beginning the first Sunday in November, 2013, the library experimented with 1:00 closings - Sundays through Thursdays. Although extended hours continued through the end of Exams, the results were compiled separately.

The experiment was prompted by a rise in the number of people in the library from 11:00 pm to midnight. It was hoped that 1:00 closings would encourage students to stay later, and might even attract late arrivals.

Additionally, the library experimented with longer hours on:

• Fridays, closing one hour later at 6:00
• Saturdays, opening an hour earlier at 10:00 and closing an hour later at 6:00.

NOTE: The trials were limited to our busiest time of the year - November / December - to ensure the highest possible counts. Had they been conducted in August / September, the results might have been half as large. Except for the occasional spike around midterm, attendance increases throughout the semester and peaks in November / December.

The results of the 1:00 experiment were significantly lower than expected with an average count of only 17, a number much lower than even 1% of our student enrollment.

The single highest count recorded was 30 and the single lowest count was 2.

• Sunday Average - 14
• Monday Average - 20
• Tuesday Average - 20
• Wednesday Average - 19
• Thursday Average - 10

Note that the determining count was made at 12:30 (as opposed to 12:45 or 1:00) to ensure the largest possible results. We know from observation that users are continually leaving from 12:30 to 1:00, and that few students arrive after 12:30. In fact, attendance peeks between 8:30 and 10:30, and then declines by 50% or more over the remaining hours.

The library extended its hours on Fridays till 6:00 p.m. beginning November 8 and running through the end of the semester. The average number of users in the building at 5:30 was 15.

There were three surprises.

• The number of the people in the building at 5:30 pm was much higher than expected.
• The number of people in the building went up between 4:00 pm and 5:30 pm instead of down.
• The number of users in the building at 5:30 during exams was only 4 - far less than during the regular semester.

The lack of consistency in the results suggests that we may need to repeat the experiment over a longer time period and look more closely at our counting methods and results. It is possible that the numbers spiked due to outside influences such as faculty committee meetings and/or the presence of tour groups iof prospective tour groups, or even by community users.

Still, the count was significantly less than even 1% our our enrollment.

The library extended its Saturday hours from 11-5 to 10-6 from November 2 till the end of the semester. Attendance was disappointing, not only during the extended hours, but throughout the day.

• The average 10:30 a.m. count was 3.
• The average 5:30 p.m. count was 5.
• The highest average count was at 4:00 and was 21.
• The average hourly count was 12.
• The lowest average count was at at 10:30 with 3 followed by 5:30 with 5.

Although we do not intend to cut Saturday hours, the unintended consequnce of the study was that attendance is so disappointing that the argument could be made that it is not worth opening the library on Saturdays until after lunch - if at all. Consider, when you look at the results, that the experiment was run during the busiest time of the year.

On an encouraging note, attendance during Exams was more than twice that of the general semester, averaging 28 vs.12 over the entire day and 11 vs. 5 at 5:30

The number of students in the building at 10:30 a.m. was 3 for both the general semester and exams. The expectation that students would take advantage of earlier Saturday hours during exams proved false.

Although not part of the extended hours experiment, the library always closely watches attendance during exam hours  - especially around opening and closing times.

Longitudinal tracking shows that attendance during exam time continues to rise with student enrollment.

Generally speaking, attendance during exams was 20% higher than that of the general semester - averaging 59 vs. 47.

Attendance during exams at 12:30 am was twice as high that of the general semester - 32 vs. 16, suggesting students don't necessarily wish to study late on a long term basis, but will do so when the need arises.

Average Counts

• 12:30 am Sundays through Thursdays - 17
• 5:30 pm Fridays - 15
• 10:30 am Saturdays - 3
• 5:30 pm Saturdays - 5
• Average half hour count during semester for all time periods - 47
• Average half hour count during exams for all time periods - 59
• Average half hour count from 10:00 through 12:30 during the regular semester by days of the week showed that Mondays (averaging 54) are our biggest night, followed by Sundays, Tuesdays, and Wednesdays (averaging 42-44), followed by Thursdays, which averaged only 18.

The experiment demonstrated that the Thrift Library's general semester hours are adequate.

The experiment demonstrated that the Thrift Library should continue to expand hours during Exams till 1:00 am, Other expansions - Fridays and Saturdays - were disappointing, though later Friday hours showed promise and may be a future possibility.

We will continue to count attendance every half hour and experiment again in the future.