Abstract
Implementing multi-item scales requires decisions as to the response format for
the scale items, which includes the optimal number of response categories and
appropriateness of item wording. We re-frame the response format issues from a
generalizability theory perspective. We examine empirical evidence from the literature and test hypotheses as to the effects of alternative response formats on
variance components and generalizability coefficients for multiple objects of
measurement.
the scale items, which includes the optimal number of response categories and
appropriateness of item wording. We re-frame the response format issues from a
generalizability theory perspective. We examine empirical evidence from the literature and test hypotheses as to the effects of alternative response formats on
variance components and generalizability coefficients for multiple objects of
measurement.
Original language | English |
---|---|
Publication status | Published - May 2004 |
Externally published | Yes |
Event | CORS/INFORMS Joint International Meeting 2004 - Canada, Banff, Canada Duration: 16 May 2004 → 19 May 2004 |
Conference
Conference | CORS/INFORMS Joint International Meeting 2004 |
---|---|
Abbreviated title | CORS/INFORMS2014 |
Country/Territory | Canada |
City | Banff |
Period | 16/05/04 → 19/05/04 |