NSSE’s Reports and Statistical Average (Mean) of Categorical Data

There is no question how importance NSSE’s (National Survey of Student Engagement) roles are in encouraging US colleges to promote continuous improvements.   Practically speaking, majority of US colleges if not all have participated in the NSSE annual survey.  NSSE in no doubt has played a significant role and it has helped college administrators around the US and accreditation agencies to identify areas where higher education institutions have the opportunity to make important improvements.

If there is one thing that is missing from all this wonderful effort is on reporting the statistical average (mean) of categorical data gathered through Likert Scale type of questionnaires.  Routinely reporting the statistical average from ordinal variables is a bit of a problem because the spacing between the values of these variables may not be the same across the levels of the variables under studied, as well as among the respondents.  NSSE is an important and influential institution which represents a leading education research.  Therefore, the way it presents its research results may have significant impacts on education policy in the country.  Furthermore, what NSSE does will significantly shape the directions of how to conduct education and social research; how to choose appropriate research methods to answer higher education questions and how to conduct hypotheses test appropriately.  Any flaw in those areas will be carried over by others because they, the researchers believe or assume what NSSE has done in its studies and research reports are the example of the best practices in the profession.

Reporting the statistical average (mean) and Standard Deviation from scale or ordinal variables, which are generated from Likert’s type of questions not only will produce bias, unintended or perhaps meaningless results, but also will lead to make incorrect statistical inference.  This mistake will be carried out and surely will impact the way the decision makers solve their institutional problem.  If colleges’ administrators recommended policy changes to the School Board or Board of Trustee based on inaccurate strategic information generated from NSSE’s reports or NSEE’s way of doing research, it will lead to even a greater waste of limited resources.

Perhaps, this is the right time for higher education industry in the US to make yet another fundamental change?

Leave a Reply

Your email address will not be published. Required fields are marked *


This site uses Akismet to reduce spam. Learn how your comment data is processed.