Glaze, Ryan M. (2009-08). The Magnitude and Extent of Malfeasance on Unproctored Internet-Based Tests of Cognitive Ability and Personality. Master's Thesis. Thesis uri icon

abstract

  • The use of unproctored internet-based testing for employee selection is widespread. Although this mode of testing has advantages over onsite testing, researchers and practitioners continue to be concerned about potential malfeasance (e.g., cheating and response distortion) under high-stakes conditions. Therefore, the primary objective of the present study was to investigate potential malfeasance effects on the scores of an unproctored internet-based cognitive ability, and a personality test. This was accomplished by implementing a within-subjects design in which test takers first completed the tests as job applicants (high-stakes) or incumbents (low-stakes) then as research participants (low-stakes). The pattern of cognitive ability test score differences was more consonant with a psychometric practice effect than a malfeasance explanation. Thus, the results suggest that, if present, there was no evidence to indicate that wide- scale or systematic malfeasance unduly affected the test scores. This may have been due to the speeded nature of the test, which was used to preempt the potential for widespread cheating. Additionally, the unproctored personality administration resulted in similar mean shifts and similar proportions of test takers being suspected of distorting their responses as that reported in the extant literature for proctored tests. In their totality, these results suggest that an unproctored internet-based administration does not uniquely threaten personality measures in terms of elevated levels of response distortion compared to proctored measures.
  • The use of unproctored internet-based testing for employee selection is
    widespread. Although this mode of testing has advantages over onsite testing,
    researchers and practitioners continue to be concerned about potential malfeasance (e.g.,
    cheating and response distortion) under high-stakes conditions. Therefore, the primary
    objective of the present study was to investigate potential malfeasance effects on the
    scores of an unproctored internet-based cognitive ability, and a personality test. This
    was accomplished by implementing a within-subjects design in which test takers first
    completed the tests as job applicants (high-stakes) or incumbents (low-stakes) then as
    research participants (low-stakes). The pattern of cognitive ability test score differences
    was more consonant with a psychometric practice effect than a malfeasance explanation.
    Thus, the results suggest that, if present, there was no evidence to indicate that wide-
    scale or systematic malfeasance unduly affected the test scores. This may have been due
    to the speeded nature of the test, which was used to preempt the potential for widespread
    cheating. Additionally, the unproctored personality administration resulted in similar
    mean shifts and similar proportions of test takers being suspected of distorting their responses as that reported in the extant literature for proctored tests. In their totality,
    these results suggest that an unproctored internet-based administration does not uniquely
    threaten personality measures in terms of elevated levels of response distortion
    compared to proctored measures.

publication date

  • August 2009