[Inclusive Research: Validation of the General Self-Efficacy Scale in Simple Language in a Sample of Students with Special Educational Needs].

BACKGROUND

Approximately 7% of pupils in Germany have special educational needs. This is accompanied by a general reduction in social participation, partly due to a reduction in language comprehension. The endeavour for social inclusion therefore includes the provision of text information in simple language. Since written surveys play an important role in social science and clinical research, this also applies to psychometric questionnaires.

OBJECTIVE

Comparison of psychometric quality of the original scale measuring self-efficacy (German version: SWE, English version: GSE) and a version of the same scale translated into simple language (SWE-LS).

METHODS

Both versions of the SWE were completed by N=94 pupils aged between 12 and 19 years with special educational needs from 7 special schools in Thuringia. Internal consistency and convergent validity of both scales were measured.

RESULTS

Both scale versions achieved satisfactory internal consistency (SWE:=0.74, SWE-LS:=0.78). However, the mean scale values differ significantly. The convergent validity was r=0.69, whereby 3 items did not correlate significantly (-0.010.21). The proportion of missing values is significantly higher for the original SWE with 16% listwise and 6.7% pairwise than for the SWE-LS with 3.2  and 0.5%, respectively.

DISCUSSION

Overall, the SWE-LS translated into simple language has a better psychometric quality than the original scale. The lower proportion of missing values also speaks for the use of the SWE-LS.

CONCLUSION

The translation of standardised questionnaires into simple language offers the possibility of also including groups of participants who have so far been excluded or who have only been surveyed indirectly (e. g. through caregivers). The cost of translation is comparable to that of translation into other languages. In the case of simple language, however, a back translation and thus a verification of the validity of the content are not possible.