In the next step, the graded response IRT model will be analyzed again for each identified domain, but without the excluded items, and the model fit statistics will be assessed. If the model has a good fit statistics, the best item for each domain will be selected according to the distribution of the threshold values across the latent scale and the discrimination value; item that has threshold values cross the whole latent scale and high discrimination value will be selected.
Sample Size Graded response IRT model requires 250 subjects to be estimated. However, approximately 500 subjects are recommended to get more accurate estimation of the model parameters [56]. We have more than 500 subject in the database that we are working on.
Phase 3: …show more content…
The experts including LBP clinicians and researchers who have experience in assessment development will be recruited from MUHC sites (Montreal General Hospital, Montreal Neurological Hospital, and Royal Victoria Hospital), Jewish General Hospital, Montreal Constance-Lethbridge Rehabilitation Centre, and Centre de Réadaptation Lucie-Bruneau. A customised letter will be sent to potential participants (Appendix 1). During the focus group, each participant will be asked to pair up with the person sitting next to him/her. Each pair will be given two copies of the best items (one in English and one in French), then, they will be asked to discuss each item in terms of the following four points: “(i) Is the wording clear and appropriate for the item? If no, how would you change it? (ii) Are the response options clear and appropriate for the item? If no, how would you change them? (iii) How difficult would it be for patients to answer the question? And (iv) Do you have any suggestions to improve the item? [57]”. At the end of the focus group, English and French versions of the assessment tool will be available for the next
Sample Size Graded response IRT model requires 250 subjects to be estimated. However, approximately 500 subjects are recommended to get more accurate estimation of the model parameters [56]. We have more than 500 subject in the database that we are working on.
Phase 3: …show more content…
The experts including LBP clinicians and researchers who have experience in assessment development will be recruited from MUHC sites (Montreal General Hospital, Montreal Neurological Hospital, and Royal Victoria Hospital), Jewish General Hospital, Montreal Constance-Lethbridge Rehabilitation Centre, and Centre de Réadaptation Lucie-Bruneau. A customised letter will be sent to potential participants (Appendix 1). During the focus group, each participant will be asked to pair up with the person sitting next to him/her. Each pair will be given two copies of the best items (one in English and one in French), then, they will be asked to discuss each item in terms of the following four points: “(i) Is the wording clear and appropriate for the item? If no, how would you change it? (ii) Are the response options clear and appropriate for the item? If no, how would you change them? (iii) How difficult would it be for patients to answer the question? And (iv) Do you have any suggestions to improve the item? [57]”. At the end of the focus group, English and French versions of the assessment tool will be available for the next