Limits...
Peer effects in unethical behavior: standing or reputation?

Pascual-Ezama D, Dunfield D, Gil-Gómez de Liaño B, Prelec D - PLoS ONE (2015)

Bottom Line: First, we show that working in the presence of peers is an effective mechanism to constrain honest/dishonest behavior compared to an isolated work situation (experiment 1).Second, we demonstrate that the mere suspicion of dishonesty from another peer is not enough to affect individual cheating behavior (experiment 2), suggesting that reputation holds great importance in a worker's self-image acting as a strong social incentives.Third, we show that when the suspicion of dishonesty increases with multiple peers behaving dishonestly, the desire to increase standing is sufficient to nudge individuals' behavior back to cheating at the same levels as isolated situations (experiment 3).

View Article: PubMed Central - PubMed

Affiliation: Department of Financial Economy and Accounting II, Universidad Complutense de Madrid, Madrid, Spain.

ABSTRACT
Recent empirical evidence shows that working in an unsupervised, isolated situation under competition, can increase dishonest behavior to achieve prestige. However, could working in a common space, in the presence of colleagues affect cheating? Here, we examine how familiar-peer influence, supervision and social incentives affect worker performance and dishonest behavior. First, we show that working in the presence of peers is an effective mechanism to constrain honest/dishonest behavior compared to an isolated work situation (experiment 1). Second, we demonstrate that the mere suspicion of dishonesty from another peer is not enough to affect individual cheating behavior (experiment 2), suggesting that reputation holds great importance in a worker's self-image acting as a strong social incentives. Third, we show that when the suspicion of dishonesty increases with multiple peers behaving dishonestly, the desire to increase standing is sufficient to nudge individuals' behavior back to cheating at the same levels as isolated situations (experiment 3).

No MeSH data available.


Efficiency in terms of time in the Ignored conditions.The average time per unit in the HS condition (where participants could not cheat) is used as a proxy for the time it would take participants to actually finish a sheet in the LS condition. Multiplying this value by the number of real sheets finished in the LS condition, we obtain a proxy for the time that it should have taken a participant to finish.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4390288&req=5

pone.0122305.g003: Efficiency in terms of time in the Ignored conditions.The average time per unit in the HS condition (where participants could not cheat) is used as a proxy for the time it would take participants to actually finish a sheet in the LS condition. Multiplying this value by the number of real sheets finished in the LS condition, we obtain a proxy for the time that it should have taken a participant to finish.

Mentions: Participants’ efficiency could be measured directly only in the HS condition. In the NS condition, it is impossible to calculate efficiency because we cannot record the number of units actually completed. In the LS condition, we know the number of units actually completed as well as the time spent by participants doing the task. However, merely dividing time by real units assumes that participants worked efficiently throughout the task and only cheated at the very end of the session by stating a greater number of units declared than those truly completed. In reality, it may be likely that dishonest participants could spend time during the task simply doing nothing. That strategy would allow dishonest workers to reduce the likelihood of being caught (or at least decrease the suspicion of cheating) by the instructor or by their peers. To get a deeper understanding of how much time might be taken up in the LS and NS conditions dishonestly idling, we have calculated a time proxy for efficient work. The average time per unit in the HS condition (where participants could not cheat) is used as a proxy for the time it would take participants in the LS condition to actually finish a sheet. Multiplying this value by the number of real sheets finished in the LS condition, we obtain a proxy for the time that it should have taken a participant to finish the whole task. That is:Efficiency_Time_Proxy=Number of sheets actually completed * Time per unit in the HS conditionThe difference between a participants´ real time spent and efficiency time proxy represents the time spent pretending to work. As we can see in Fig 3, the average difference in the four conditions (I, FP, L and TL) was 4.78, 7.18, 9.88 and 5.03 minutes, respectively. In a one-way ANOVA on those differences with Presence of Others as a factor, we found a main effect [F(3,88) = 6.69; p = .000; μ2 = .18], pointing out significant differences between L and I (p = .001), and between L and TL (p = .002). No differences were found between L and FP (p = .191). Together, these results show that in TL, participants not only cheated more (the proportion of finished sheets compared to those declared was the smallest), but they also spent more time pretending to work.


Peer effects in unethical behavior: standing or reputation?

Pascual-Ezama D, Dunfield D, Gil-Gómez de Liaño B, Prelec D - PLoS ONE (2015)

Efficiency in terms of time in the Ignored conditions.The average time per unit in the HS condition (where participants could not cheat) is used as a proxy for the time it would take participants to actually finish a sheet in the LS condition. Multiplying this value by the number of real sheets finished in the LS condition, we obtain a proxy for the time that it should have taken a participant to finish.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4390288&req=5

pone.0122305.g003: Efficiency in terms of time in the Ignored conditions.The average time per unit in the HS condition (where participants could not cheat) is used as a proxy for the time it would take participants to actually finish a sheet in the LS condition. Multiplying this value by the number of real sheets finished in the LS condition, we obtain a proxy for the time that it should have taken a participant to finish.
Mentions: Participants’ efficiency could be measured directly only in the HS condition. In the NS condition, it is impossible to calculate efficiency because we cannot record the number of units actually completed. In the LS condition, we know the number of units actually completed as well as the time spent by participants doing the task. However, merely dividing time by real units assumes that participants worked efficiently throughout the task and only cheated at the very end of the session by stating a greater number of units declared than those truly completed. In reality, it may be likely that dishonest participants could spend time during the task simply doing nothing. That strategy would allow dishonest workers to reduce the likelihood of being caught (or at least decrease the suspicion of cheating) by the instructor or by their peers. To get a deeper understanding of how much time might be taken up in the LS and NS conditions dishonestly idling, we have calculated a time proxy for efficient work. The average time per unit in the HS condition (where participants could not cheat) is used as a proxy for the time it would take participants in the LS condition to actually finish a sheet. Multiplying this value by the number of real sheets finished in the LS condition, we obtain a proxy for the time that it should have taken a participant to finish the whole task. That is:Efficiency_Time_Proxy=Number of sheets actually completed * Time per unit in the HS conditionThe difference between a participants´ real time spent and efficiency time proxy represents the time spent pretending to work. As we can see in Fig 3, the average difference in the four conditions (I, FP, L and TL) was 4.78, 7.18, 9.88 and 5.03 minutes, respectively. In a one-way ANOVA on those differences with Presence of Others as a factor, we found a main effect [F(3,88) = 6.69; p = .000; μ2 = .18], pointing out significant differences between L and I (p = .001), and between L and TL (p = .002). No differences were found between L and FP (p = .191). Together, these results show that in TL, participants not only cheated more (the proportion of finished sheets compared to those declared was the smallest), but they also spent more time pretending to work.

Bottom Line: First, we show that working in the presence of peers is an effective mechanism to constrain honest/dishonest behavior compared to an isolated work situation (experiment 1).Second, we demonstrate that the mere suspicion of dishonesty from another peer is not enough to affect individual cheating behavior (experiment 2), suggesting that reputation holds great importance in a worker's self-image acting as a strong social incentives.Third, we show that when the suspicion of dishonesty increases with multiple peers behaving dishonestly, the desire to increase standing is sufficient to nudge individuals' behavior back to cheating at the same levels as isolated situations (experiment 3).

View Article: PubMed Central - PubMed

Affiliation: Department of Financial Economy and Accounting II, Universidad Complutense de Madrid, Madrid, Spain.

ABSTRACT
Recent empirical evidence shows that working in an unsupervised, isolated situation under competition, can increase dishonest behavior to achieve prestige. However, could working in a common space, in the presence of colleagues affect cheating? Here, we examine how familiar-peer influence, supervision and social incentives affect worker performance and dishonest behavior. First, we show that working in the presence of peers is an effective mechanism to constrain honest/dishonest behavior compared to an isolated work situation (experiment 1). Second, we demonstrate that the mere suspicion of dishonesty from another peer is not enough to affect individual cheating behavior (experiment 2), suggesting that reputation holds great importance in a worker's self-image acting as a strong social incentives. Third, we show that when the suspicion of dishonesty increases with multiple peers behaving dishonestly, the desire to increase standing is sufficient to nudge individuals' behavior back to cheating at the same levels as isolated situations (experiment 3).

No MeSH data available.