The wording of a questionnaire may appear clear and precise, yet convey different meanings to different respondents. Although the purpose of Rasch analysis is to construct generalizable measures along the variable of interest, the measures by themselves have no value if it is unclear what is being measured.
Bookstein and Lindsay (B&L, 1989) investigate this problem by designing a questionnaire to discover what respondents understand by a question frequently asked in their specialty of Library Science. The question is "How often do you use a library?" The answers received to this question from the general public, library clients and other relevant audiences are crucial to assignment of resources in libraries and to the livelihood of librarians.
But what is "library use"? B&L formulate 15 probes into this concept, each starting with "Would you say you used a library today if ... ?" They administered these probes to 33 students, "S", at a Library Graduate School, 4 librarians, "L", and 5 non-librarians, "N".
Their results are summarized in the Figure. The 0-to-10 logit scale measures strictness of interpretation with 10 being most strict and 0 most loose. Each "L", "N" or "S" indicates one person. Placement in the Figure indicates that the person tends to agree that activities listed above are library users, but activities listed below are not. The lowest "N" responded that all activities are library uses. Librarians are located on the lower, loose side in interpreting "library use" (as it is in their own interest to be!). Non-librarians exhibit a "black or white" interpretation, being towards one or other end of the person distribution. Students' interpretations cover the spectrum.
Everyone agreed that "library use" includes the "major" traditional library-centered activities of reading articles, referring to card catalogs, conducting online searches and checking out books. Below this, three strata can be identified, based on how intensively library-specific resources are employed. "Minor" use is when no intentional reading of library material occurs. "Petty" use covers other activities in which, at best, only incidental benefit is obtained from traditional library material. "Building" use is that for which any convenient public building would suffice.
This study illustrates how important shared understanding is to the construction of measures and, hence, knowledge. A simple question, such as "How often do you use a library?", has a spectrum of different meanings to different respondents. Since complete commonality of understanding is unobtainable, another attribute of sound measurement, replication, must be its substitute. Through asking variations on the same question, in different ways and with different intensities, the answer to the researcher's intended query can be constructed step-by-step.
Abraham Bookstein 1994 RMT 8:1 P.333
Bookstein A & Lindsay A (1989) Questionnaire ambiguity: a Rasch scaling model analysis. Library Trends 38:2 p. 215-236.
Just what do you mean ...? Bookstein A. … Rasch Measurement Transactions, 1994, 8:1 p.333
The URL of this page is www.rasch.org/rmt/rmt81a.htm
Website: www.rasch.org/rmt/contents.htm