Standardizations in methods of data collection: Adverse effects on reliability?

Abstract

In psychological methods of data collection restrictions set by the investigator quite often diminish the freedom of decision of the subjects: the subjects can only act within the frame of reference provided by the researcher. This paper wants to show the disadvantageous influence such predecisions can have on the results of an inquiry if they do not correspond to the inner representations of the subjects. This is demonstrated empirically for the reliability of data: the results show that the reliability of rating tasks decreases when constructs and the number of points in a rating scale are provided. The lack of a remainder category also tends to diminish the reliability. The results can be interpreted as an argument for person-centred data collection, which either dispenses with providings laid down by the investigator or includes providings being adjusted to the inner representations of the subjects.