Construction and validation of a psychometric tool for generational identity

There is a lack of a psychometric tool for generational identity. We have conducted two studies involving senior Bangladeshi nationals, containing both independence war veterans and non-veterans, to develop a new generational identity scale (GIS). The first study (N=300) prepared an initial pool of 31 items and got them vetted by expert judges, which retained 21 items to form the provisional GIS (GIS-21). An exploratory factor analysis on GIS-21 excluded nine items and offered a two-factor solution: (i) identification with the generation and (ii) awareness of the generational importance. The second study (N=176) ran a confirmatory factor analysis on the resulting GIS-12 and dropped another item to achieve a better model fit (χ2/df =74.74/42 =1.78, CFI =0.96, TLI =0.95, RMSEA =0.05, and SRMR =0.05). The remaining 11-item GIS (GIS-11) showed excellent reliability (mean inter-item correlation coefficient =.44, Cronbach’s alpha =.89) and satisfactory temporal stability (ICC =.61, 95% CI =0.30 to 0.78) over a 4-week interval. The Bangladesh independence war veterans scored higher on GIS-11 than did the non-veterans, indicating the scale’s diagnostic validity. The scale’s moderate correlation (rs=.63, p