Priming Public Managers to Participate in Research: Evidence from a Low-Intensity Survey Experiment

How to cite

Pedersen, Mogens Jin; Petersen, Niels Bjørn Grund; Pedersen, Rasmus Tue. 2020. "Priming Public Managers to Participate in Research: Evidence from a Low-Intensity Survey Experiment." AEA RCT Registry. AEARCTR-0006526. https://doi.org/10.1257/rct.6526-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Mogens Jin Pedersen* , Niels Bjørn Grund Petersen , Rasmus Tue Pedersen

* Primary investigator

Registry ID
AEARCTR-0006526
DOI
https://doi.org/10.1257/rct.6526-1.0
First registered
Last updated
Status
On going
Trial dates
2020-09-23 – 2020-10-07
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/6526

Abstract

Much public management research uses data involving voluntary participation from public managers in surveys or interviews. To what extent may low-intensity text appeals promote public managers’ willingness to volunteer as respondents in research studies? Are particular types of text appeal more effective than others are? Using a between-subjects survey experimental design, this study examines the effects of various text appeals on public managers’ willingness to participate in a qualitative research interview. At the end of an electronic survey sent to about 10,000 public managers in Denmark, respondents are invited to participate in a follow-up interview study. Each respondent is randomly assigned to either a control group (basic text) or one of two treatment groups, TG1 or TG2. Respondents in TG1 are exposed to brief text serving “controlled extrinsic” forms of motivation, whereas respondents in TG2 are exposed to brief text appealing to prosocial forms of motivation. We examine and compare the effects of the treatments on the respondents’ survey-reported willingness to participate in the follow-up interview study.

Intervention

The interventions consist of low-intensity text appeals appearing at the end of an electronic survey. All survey respondents are notified that a set of qualitative interviews are to be conducted following the survey examination. By random assignment, respondents are then exposed to one of the following five text appeals:

(1) “We want to hear if you are willing to participate.” [in Danish: “I den forbindelse vil vi høre, om du vil deltage.”]
(2) “With your participation, you will contribute with your special knowledge and unique insights and experience” [in Danish: “Med din deltagelse vil du bidrage med din særlige viden og unikke oplevelser og erfaringer.”]
(3) “With your participation, you will have the chance to further reflect upon you own leadership style and development” [in Danish: “Med din deltagelse vil du få mulighed for at reflektere nærmere over din egen ledelsesstil og lederudvikling.”]
(4) “With your participation, you will help establish new and important knowledge about management training programmes in Denmark – which ultimately will be of benefit to citizens and society.” [in Danish: “Med din deltagelse vil du bidrage til at skabe ny og vigtig viden om lederuddannelser i Danmark – hvilket i sidste ende vil være til stor gavn for borgere og samfundet.”]
(5) “With your participation, you will help establish new and important knowledge about management training programmes in Denmark – which ultimately will be of benefit to other managers.” [in Danish: “Med din deltagelse vil du bidrage til at skabe ny og vigtig viden om lederuddannelser i Danmark – hvilket i sidste ende vil være til stor gavn for andre ledere.”]

Experimental design

About 10,000 public managers in Denmark receive an invitation to participate in an electronic survey.
At the end of the survey, respondents are invited to participate in a follow-up interview study. In particular, all respondents are exposed to a survey item stating that a set of qualitative interviews are to be conducted following the survey examination. Our experimental design involves manipulation of the exact survey item text.

Each respondent is randomly assigned to either a control group (CG) or one of two treatment groups, TG1 or TG2. Respondents are assigned to CG, TG1, and TG2 with equal probability (i.e., 33/33/33 spilt).

CG respondents receive the following text: “We want to hear if you are willing to participate.”

TG1 respondents receive one of the following two texts (assigned at random) serving “controlled extrinsic” forms of motivation:

(a) “With your participation, you will contribute with your special knowledge and unique insights and experience”
(b) “With your participation, you will have the chance to further reflect upon you own leadership style and development”

TG2 respondents receive one of the following two texts (assigned at random) appealing to prosocial forms of motivation:

(a) “With your participation, you will help establish new and important knowledge about management training programmes in Denmark – which ultimately will be of benefit to citizens and society”
(b) “With your participation, you will help establish new and important knowledge about management training programmes in Denmark – which ultimately will be of benefit to other managers.”

As our outcome measure, all respondents are asked to report whether they may be contacted for an interview. Response categories are “yes” and “no.” Respondents who agree to participate are asked to report their contact email address (open textbox).

We test the following hypotheses:

H1: Text appeals serving “controlled extrinsic” forms of motivation increase willingness to participate in research (relative to a basic text)
H2: Text appeals serving prosocial forms of motivation increase willingness to participate in research (relative to a basic text)
H3: Text appeals serving prosocial forms of motivation increase willingness to participate in research (relative to text appeals serving “controlled extrinsic” forms of motivation)

Design details

Randomization method
Randomization is carried out by simple randomization by computer (randomization based on a single sequence of random assignments).
Randomization unit
The individual survey respondent.
Clusters
Providing an exact estimate is difficult. Our observations will be those out of the 10,000 public managers who are invited to participate in the survey who (a) chose to participate and (b) reach the end of the survey (where are survey item interventions occur). Given uncertainty about the survey response rates and survey dropout attrition, our conservative estimate is around 1,800 observations
Observations
Providing an exact estimate is difficult. Our observations will be those out of the 10,000 public managers who are invited to participate in the survey who (a) chose to participate and (b) reach the end of the survey (where are survey item interventions occur). Given uncertainty about the survey response rates and survey dropout attrition, our conservative estimate is around 1,800 observations
Arms
SAMPLE SIZE (OR NUMBER OF CLUSTERS) BY TREATMENT ARMS CG: 600 public managers T1: 600 public managers (300 to text TG1a; 300 to text TG1b) T2: 600 public managers (300 to text TG2a; 300 to text TG2b)
Primary outcomes
Willingness to participate in a follow-up interview study (1) or not (0).

Cited by

No citation data available for this registration yet.

Keywords