Using Average Scores on A Likert Scale

surveys
analysis

#1

Greetings,

I am reviewing project OUTCOME for my library system and had a general question about the analysis I am seeing. There is debate in the research whether considering a Likert scale (i.e. Strongly Disagree to Strongly Agree) as continuous data is appropriate and valid. I tend to agree that Likert scales are not continuous and should be considered ordinal data. This would preclude the use of averages as a valid descriptive statistic which seems to be the majority of the analysis I am finding. This also brings additional ambiguity around the appropriate value to assign to the neutral response options (i.e. Neither Agree nor Disagree).

I would like to better understand project OUTCOME’s thoughts on these topics.

Thank you!
Garrett


#2

Hi Garrett,

Good question. When the data dashboard developers created the tool, the decision was made to use averages in some areas to facilitate comparisons between groups. However, between the reports and different dashboards, users have a variety of other options for viewing and analyzing their data, including by count (Detail dashboard) and stacked bar charts and percentages (reports & Detail dashboard), in addition to weighted average score (reports, several dashboards). Users can also download their raw data as a csv file.

I’m the chair of PLA’s Measurement, Evaluation, and Assessment Committee, which oversees Project Outcome, and we do plan to review how statistics are reported and consider whether any changes need to be made to the dashboard.

Thanks for bringing this up!
Linda Hofschire


#3

Linda,

Thank you for your detail response. It is appreciated.

I actually have another question. Summer Reading has a survey with pre-built questions. It is my understanding, and noted by PEW, that children under 18 (in
most cases) cannot legally consent to a survey. Are you aware of how other libraries are addressing this?


#4

Hi Garrett,

Thank you for the great questions. In 2016, the Task Force decided to split the Summer Reading topic into 3 categories based on audience type: Adult, Teen/Child, and Caregiver. We know some libraries may have policies against surveying minors, so we have the Caregiver survey to capture that data instead. The Summer Reading Caregiver surveys the parents or caregivers to measure if they believe their child maintained or increased their reading skills, is a more confident reader, reads more often, and uses the library more often. Please let me know if you have any other questions!