Psychometric Scale Construction - Item Response Theory and SEM.
- or -
Post a project like this$$$
- Posted:
- Proposals: 1
- Remote
- #1747929
- Expired
Description
Experience Level: Expert
Estimated project duration: less than 1 week
The company for whom I work is investigating the possibility of expanding an existing measure of personality – ie. increasing the number of items per scale from 10 to anywhere between 10 and 20. We would like to do this with the highest level of rigour possible, as we appreciate the slightly unusual nature of adding items to a well validated scale (instead of trying to remove them, as it typical).
We would therefore like to approach this using Item Response Theory and then confirm the model through Structural Equation Modelling.
The data is already cleaned and coded (in groups of 20 items per sub-scale, with 1-10 being the original items). We have a total of 500 respondents – whilst we appreciate this would be a relatively small sample size for a total of number of 600 items, our a priori knowledge of 30 validated sub-scales within this pool means that we would treat each block of 20 as its own scale. Therefore we would look to examine unidimensionality and respective item loadings within each sub-scale (of 20 items) independently from each other.
The process we would like to see adopted for this analysis is as follows:
1. Application of a continuous Item Response Model to each sub-set of 20 items – eg. Samejima’s Graded Response.
2. Retention of items that load well onto a primary factor for each sub-set of 20 items.
3. Confirmation of Model fit (reporting of Root Mean Square Error of Approximation, Comparative Fit Index, and Standardised Root Mean Square Residual.) and Item fit (Bivariate residual tests of Item fit).
The data is currently in .xlsx and .sav format, so can be provided in either.
We would like a clearly presented output file, and your interpretation of which items should be retained for each scale with the respective statistics confirming model fit. We would also like a brief outline of the specific analyses that you ran (this can be short bullet points, but sufficiently detailed so that your method could be replicated).
Tom C.
0% (0)Projects Completed
-
Freelancers worked with
-
Projects awarded
0%
Last project
17 Apr 2024
United Kingdom
New Proposal
Login to your account and send a proposal now to get this project.
Log inClarification Board Ask a Question
-
Hello.
Could you provide the xlsx at hand?
Thanks.
520936
We collect cookies to enable the proper functioning and security of our website, and to enhance your experience. By clicking on 'Accept All Cookies', you consent to the use of these cookies. You can change your 'Cookies Settings' at any time. For more information, please read ourCookie Policy
Cookie Settings
Accept All Cookies