The Performance Effects of Relative Performance Information and Feedback Detail: Evidence from a Field Experiment in Austria

Last registered on July 01, 2019

Pre-Trial

General Information

Title

RCT ID

AEARCTR-0004141

Initial registration date

May 02, 2019

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

May 23, 2019, 5:43 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated

July 01, 2019, 4:23 AM EDT

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Country

Austria

Region

Styria

Primary Investigator

Name

Sebastian Koppers

Affiliation

WHU – Otto Beisheim School of Management

Contact Primary Investigator

Other Primary Investigator(s)

Additional Trial Information

Status

On going

Start date

2018-10-09

End date

2019-08-31

Keywords

Firms & Productivity

Additional Keywords

Performance measurement, feedback

JEL code(s)

D83, D91, L83, M12, M41, M53, M54, Z20, Z22

Secondary IDs

Abstract

We seek to establish how feedback can be customized to optimize performance. We will extend knowledge on this topic by using a field experiment in a competitive sport setting (a commercial training simulator used by elite football clubs) to scientifically establish the types of feedback that work best on average and for players of different profiles. Specifically, we will test the performance effects of relative performance feedback (absolute vs relative vs absolute and relative) and feedback detail (high-level vs detailed). We will do so by randomly assigning different types of feedback to the players (high-level and absolute, high-level and relative, high-level and both absolute and relative, detailed and absolute, detailed and relative, detailed and both absolute and relative).

External Link(s)

Registration Citation

Citation

Koppers, Sebastian. 2019. "The Performance Effects of Relative Performance Information and Feedback Detail: Evidence from a Field Experiment in Austria." AEA RCT Registry. July 01. https://doi.org/10.1257/rct.4141-3.0

Former Citation

Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Experimental Details

Interventions

Intervention(s)

Subjects perform a football (soccer) drill in a training simulator at a field site (professional football training facility). Between each round of the drill, the subjects receive feedback on their performance on that round of the drill.

Intervention Start Date

2018-10-09

Intervention End Date

2019-08-31

Primary Outcomes

Primary Outcomes (end points)

We employ two measures of player performance on the drill: accuracy of passes (as measured by meters from target), and reaction time of player (as measured by duration in time in seconds from release of ball until pass hits target area). These measures are captured by the football training simulating. We also measure the subjects' heart rates whilst they perform the drill. This serve to proxy for effort (level of heart rate) and stress (variability of heart rate).

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

We conduct a post-experiment survey to measure the following constructs: perceived value of feedback about performance, social comparison involvement, information quantity, focus, shift of focus, focus on best or worst, influence of feedback on focus, motivation, trust, and perceived threat of feedback.

Secondary Outcomes (explanation)

Perceived value of feedback is based on items from Zingoni & Byron (2017). Social comparison involvement is based on Tafkov (2013). Focus, shift of focus, focus on best or worst, influence of feedback on focus, information quantity and motivation are measured using items specifically developed for this study. The measure of trust is based on Kizilcec (2016). Perceived threat of feedback uses the instrument developed in Zingoni & Byron (2017).

Experimental Design

We employ a fully-interacted 3x2 design to evaluate the performance effects of relative performance feedback and feedback detail.

Experimental Design Details

We employ a fully-interacted 3x2 design to evaluate the performance effects of relative performance feedback and feedback detail. In the relative performance feedback treatment arm of the experiment, subjects are randomly assigned to one of three conditions: relative feedback (subjects see their performance rank on each of the performance measures), absolute feedback (subjects see their absolute performance on each of the performance measures), both absolute and relative feedback (subjects see both their performance rank and absolute performance on each of the performance measures). In the feedback detail treatment arm of the experiment, subjects are randomly assigned to one of two conditions: high-level (subjects see their aggregate performance on the performance measures) and detailed (subjects see their disaggregated performance on the performance measures, broken out by different features of the training drill).

Randomization Method

Subjects are randomly assigned to one of six treatment groups. To do this, we employ a quasi-blocked randomization design, whereby we generate a random sequence of numbers to assign male subjects to treatment and a random sequence of numbers to assign female participants to treatment. We do this to make sure that fewer female participants do not cluster in one treatment group and bias experimental results (true blocked randomization not possible due to no-shows of registered participants).

Randomization Unit

Individual

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

180 individuals

Sample size: planned number of observations

180 individuals

Sample size (or number of clusters) by treatment arms

We assigned our total sample evenly across each of the treatment cells: Absolute & High- level (30 subjects), Absolute & Detailed (30 subjects), Relative & High-level (30 subjects), Relative & Detailed (30 subjects), Absolute & Relative & High-level (30 subjects), Absolute & Relative & Detailed (30 subjects)

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Performance as measured by handling time. Given a sample size of 90 subjects per treatment arm, the minimum detectable effect size is .08 seconds, which is equivalent to change in performance by 0.4 standard deviations (or 3% of the mean level of performance).

Supporting Documents and Materials

IRB