Page 101 - Medical students’ self-regulated learning in clinical contexts

Analysis
We analyzed the data using dedicated so ware (PQMethod 2.11) to perform a by- person factor analysis to extract the sta s cal factors, corresponding to pa erns in behavior.28 This technique clusters individuals with similar answers together, rather than items, as is the case in tradi onal factor analysis. We conducted the analysis using common techniques for Q-methodology (centroid factor rota on and varimax rota on).20 We iden ed all factor structures supported by the data by using the commonly used criteria of Eigenvalue >1.00, and a minimum of two par cipants who were associated sta s cally signi cantly (p<0.05) with the factor.20 Three researchers (JB, PT, EH) studied and discussed the factor structures supported by the data to see whether the factors iden ed in each solu on represented coherent pa erns in SRL behavior. The nal factor solu on was selected because it consisted of ve clearly interpretable and dis nct factors that together portrayed a comprehensive summary of the (quan ta ve and qualita ve) data.
Next, an idealized sort was computed for each of the retained factors. This is a weighted ranking of the statements for the students associated with the factor, based on their rankings of the statements and their correla on coe cients with the factor as weights. This idealized sort represents how a hypothe cal student with ex- actly that pa ern in SRL behavior would have ranked the statements, and serves as the basis for interpreta on and descrip on of the results. From the idealized sorts, dis nguishing statements between factors and consensus statements across factors were iden ed. Dis nguishing statements signi cantly di er in posi on in the ide- alized sort for a certain factor compared to all other factors, and consensus state- ments do not have a signi cantly di erent posi on for any factor. JB, PT and EH then interpreted all factors by using the idealized sorts, the dis nguishing and consensus statements, and the narra ve data from the par cipants signi cantly associated with each factor. Lastly, we provided a descrip ve label to all factors for the ease of re- membrance and future use.
Ethics
The ethical review board of the Netherlands Associa on for Medical Educa on (NVMO) approved the study, le number 376. All par cipants provided informed consent.
99
Chapter 5 Pa erns in clinical students’ self-regulated learning behavior: a Q-methodology study