Plagiarism A behavioural economics aproach

Document Sample
Plagiarism A behavioural economics aproach Powered By Docstoc
					         Rachel McCloy & Patricia Riddell
Dept of Psychology , University of Reading

     With Emily Hancock and Jonathan Haenen
 Plagiarism is an increasingly large problem in UK
 academic settings, especially with increasing use of the
 internet (Szabo & Underwood, 2004).

 Its detection is time consuming for academic staff, and
 it has a cost for students, who may face penalties for
 plagiarism in their academic work (whether
 intentional or unintentional: Evans, 2004).
Behavioural Economics
 Aims to go beyond the purely “rational” model of
  human decision makers common to many normative
  theories in economics
 Takes into account other factors
   E.g., context, emotional state and arousal, anticipated
   Arousal and decision making (Ariely & Loewenstein,
Behavioural economics and
 Has been previous research into some related
   Cheating (e.g., Mazar & Ariely, 2006; Mazar, Amir &
    Ariely, 2008; Nagin & Pogarsky, 2003; Vohs & Schooler,

   Honour codes (e.g., (McCabe & Trevino, 1993)
 Maxar, Air & Airley (2008)
 General knowledge or Math questions
 Participants paid by results:
    Experimenter scored
    Participant scored
    Participant scored & paper shredded
    Participant scored, paper shredded and pays self
 Participants given the opportunity to cheat did so.
 But: note, still only a little cheating – couple of

 Cheating increased if payment was in tokens that
  would later be exchanged for money, rather than in
  money directly
Cheating: What decreases the
likelihood of cheating?
 Increasing chances of detection (see also Nagin &
 Pogarsky, 2003)

 Cueing moral codes
    Ten commandments (Mazar,Amir & Ariely, 20o8)
    Honour codes (Mazar, Amir & Ariely, 2008; McCabe &
     Trevino, 1993)
    Free will (Vohs & Schooler, 2008)
 Increasing likelihood of detection will decrease

 Cueing relevant norms will reduce plagiarism:
    Honesty?
    Originality?
    Ownership?

 How do students perceive plagiarism?
   Extent of dishonesty can reflect whether or not people
    can justify their actions (e.g., coke vs. money)
Our project
 Funded by HEA Psychology network

 Studies (work in progress)
    Practical studies
       Changing tutorial procedures for our students

   Experimental studies
       Examining the effects of different factors on
Study 1
 Two main questions:
   1. Does increasing the (perceived) chance of detection
      decrease plagiarism?

  2.   Do students misrepresent what counts as plagiarism?

 Took a baseline of levels of plagiarism by submitting
   all essays from the previous 2nd year cohort to
      Study 1 – Changing Tutorial Procedures
All our second year students were be told, before their first
essay was due, that all essays this year will be submitted to
TurnItIn (increasing chances of detection)

When the first essays are submitted, all essays were assessed
using TurnItIn. We predicted a reduction in plagiarism in
comparison to baseline.

Students were then given their TurnItIn report for their first
essay before they wrote their second essay.

We expected a further reduction in plagiarism when students
have a more personal idea of what this is.
Study 1
 Participants
    Part 2 psychology students at University of Reading

   Baseline 501 student essays from last year’s Part 2
   Essay 1 – 128 essays
   Essay 2 – 126 essays
Results – Study 1
Results – Study 1
 No significant difference between Baseline and Essay 1
   Increasing the chances of detection did not have a
    significant effect on levels of plagiarism

 Significantly lower levels of plagiarism on Essay 2 than
 at Baseline and Essay 1
   Suggests that students lack understanding of what
    counts as plagiarism
     Study 1 – Effect of Priming on Plagiarism
                 Study in progress

Participants (50 participants in each primed group and a
control group with no priming) are told that we are
investigating the links between different language skills and
will be asked to complete two tasks.

Task 1: Scrambled sentence task (priming )

Task 2: Summary task

Tested 51 participants to date
 Lie not did sheep he

In words insect own her
     Study 1 – Effect of Priming on Plagiarism
Control group: 30 neutral sentences

Honesty group: 15 neutral sentences, 15 sentences priming

Originality group: 15 neutral sentences, 15 sentences priming

Task 2 requires participants to summarise two short texts .
Study 2 – preliminary results
                  Only 17 participants in
                   each condition so far

                  No significant differences

                  Trend across all measures
                   that we’ve taken so far that
                   originality primes lead to
                   lower levels of plagiarism
                   than honesty and control
Next steps
 Finish study 2

 Study 3 – changing coversheets
    Neutral, honour code, originality pledge

 Study 4 – priming ownership
    Name only, basic ownership cue, rich ownership cue
 May be possible to apply insights from behavioural
  economics to reducing plagiarism

 However, manipulations that have been effective in
  reducing cheating do not directly map across

 Students do not see low-level plagiarism as cheating
    Most effective manipulation highlighted low-level plagiarism

 Originality and ownership cues may prove more effective
  than honour codes

Shared By: