Discussion: View Thread

  • 1.  Extreme response set

    Posted 01-15-2009 12:14

    I am trying to locate a literature on a specific response set with respect to Likert scales. A respondent might use an extreme rating (1 or 7 on a 7 point scale) because they she has a strongly held opinion.  However, someone with the identical opinion might give a less extreme rating (2 or 6 on a 7 point scale).  Thus, separate from the expression of a true score and random measurement error, the score on a Likert scale may reflect individual differences in a response set. The response preference for extreme vs more moderate ratings probably goes by more than one name.  Terms that I have found for this effect are "extreme response set" and "expression response bias."    I seek assistance because these search terms yield a few articles that are decades old.

    I want to know the following:
    1) What terms are used for this response set?
    2) Are there mean demographic differences (age, sex, race) in preference for extreme vs moderate responses?
    3) Are there personality or cognitive ability correlates of such differences?

    Thanks,

    Mike
    Michael A. McDaniel, Ph.D.
    Professor - Human Resources and
         Organizational Behavior
    Department of Management,
    Research Professor of Psychology
    Virginia Commonwealth University
    1015 Floyd Ave, PO Box 844000
    Richmond, VA 23284-4000


    http://www.people.vcu.edu/~mamcdani/
    voice: 804.827.0209
    e-mail:    MAMcDani@vcu.edu
    skype: MichaelAMcDaniel

    		Located in Richmond, VA, the VCU School of Business is the home of the Center for the Advancement of Research Methods and Analysis.



  • 2.  Extreme response set

    Posted 01-15-2009 15:13
    Michael,

    I am not sure it is reflective of what you are looking for, but the
    terms 'severity bias' (meaning low end of the scale), 'central
    tendency bias' (the central/middle of the scale) and 'leniency bias'
    (high end of scale) may also apply to your work. The difference
    seems to be a specification of a particular part of the scale as
    opposed to extremity at both ends.

    Chris

    >>> Michael A McDaniel/AC/VCU <mamcdani@VCU.EDU> 1/15/2009 12:13 PM
    >>>
    I am trying to locate a literature on a specific response set with
    respect
    to Likert scales. A respondent might use an extreme rating (1 or 7
    on a 7
    point scale) because they she has a strongly held opinion.
    However,
    someone with the identical opinion might give a less extreme rating
    (2 or
    6 on a 7 point scale). Thus, separate from the expression of a
    true score
    and random measurement error, the score on a Likert scale may
    reflect
    individual differences in a response set. The response preference
    for
    extreme vs more moderate ratings probably goes by more than one
    name.
    Terms that I have found for this effect are ?extreme response set?
    and
    ?expression response bias.? I seek assistance because these
    search
    terms yield a few articles that are decades old.

    I want to know the following:
    1) What terms are used for this response set?
    2) Are there mean demographic differences (age, sex, race) in
    preference
    for extreme vs moderate responses?
    3) Are there personality or cognitive ability correlates of such
    differences?

    Thanks,

    Mike

    Michael A. McDaniel, Ph.D.
    Professor - Human Resources and
    Organizational Behavior
    Department of Management,
    Research Professor of Psychology
    Virginia Commonwealth University
    1015 Floyd Ave, PO Box 844000
    Richmond, VA 23284-4000

    http://www.people.vcu.edu/~mamcdani/
    voice: 804.827.0209
    e-mail:MAMcDani@vcu.edu
    skype: MichaelAMcDaniel
    Located in Richmond, VA, the VCU School of Business is the home of
    the
    Center for the Advancement of Research Methods and Analysis.