Understanding the impact of Fatwas on health behavior: Evidence from multiple online experiments in Indonesia

Last registered on January 03, 2023


Trial Information

General Information

Understanding the impact of Fatwas on health behavior: Evidence from multiple online experiments in Indonesia
Initial registration date
December 22, 2022

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
January 03, 2023, 5:12 PM EST

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



Primary Investigator

Bernhard Nocht Institute for Tropical Medicine (BNITM)

Other Primary Investigator(s)

PI Affiliation
Universitas Indonesia
PI Affiliation
University of Goettingen

Additional Trial Information

On going
Start date
End date
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Fatwas, a ruling on a point of Islamic law given by a recognized authority, are important moral guidelines for Muslims around the world. While economists acknowledge that religious rules can influence individual and collective behaviors, there is still little rigorous evidence on the impact of religious rulings on health behavior. More specifically, the cognitive processes underlying the link between Islamic religious ruling and health behavior is not yet well understood.
Our study context is Indonesia – the country with the largest Muslim population globally. With over one third of the population smoking, smoking still represents an important public health concern. In 2010, one of the country’s most important Islamic organization – Muhammadiyah –issued to an anti-smoking fatwa in order to address this issue. In this study we explore the effects of this ruling. We employ five distinct online experiments that involve a total of 14,500 individuals, in order to better understand an individual’s cognitive processes related to smoking. More specifically, we investigate (i) the role of information uptake and avoidance, (ii) motivated reasoning and storing of information, and (iii) cognitive strategies in dealing with challenging information.
External Link(s)

Registration Citation

Dartanto, Teguh, Renate Hartwig and Jan Priebe. 2023. "Understanding the impact of Fatwas on health behavior: Evidence from multiple online experiments in Indonesia." AEA RCT Registry. January 03. https://doi.org/10.1257/rct.10690-1.0
Sponsors & Partners

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

Request Information
Experimental Details


The interventions consist of the following:
- Menu of different article titles to chose from
- Menu of different articles to read
- Psychological primes
- Cognitive load tasks
- Fact-checking tasks
Intervention Start Date
Intervention End Date

Primary Outcomes

Primary Outcomes (end points)
Preferences/Decisions in terms of (i) article selection and (ii) information processing
Primary Outcomes (explanation)
I) Article selection: Our dependent variables relate to whether subjects have a preference for confirmatory articles or not. We will look at three variables namely whether (i) individuals select articles that are from the religious body in line with their religious affiliation (binary), (ii) individuals select articles that are in line with their own smoking practice (binary), and (iii) individuals select articles that are reflecting their own religious identity and their own smoking practice (ordinal).

II) Information processing: We measure information processing in two ways: (i) by the title as a binary measure of attentiveness which reflects whether the subject can recall the correct title of the article that they have read, and (ii) measures on the recall of the content. Our recall measures on the content relate to the number of arguments subjects can correctly recall from the shown article (count data) and whether subjects can correctly remember at least three arguments as listed in the paragraphs (binary indicator).

Secondary Outcomes

Secondary Outcomes (end points)
Secondary outcomes relate to variables that present mechanisms.
Secondary Outcomes (explanation)
: Secondary outcomes include variables that we will use to investigate mechanisms. Secondary outcomes relate to (a) respondent characteristics (socio-demographics, cognitive skills, personality traits, economic preferences, religiousness, health status, institutional trust, smoking behavior), (b) respondent views on provided article (rating of article, willingness to share the article), (c) impact of the article on policy preferences (smoking regulations, attitudes towards other topics possibly considered controversial in Islam), and (d) survey environment (pressure, relevance).

Experimental Design

Experimental Design
The study applies 5 online experiments. As part of the experiments online subjects are randomized into different treatment groups. In 4 out 5 experiments subjects are randomized in terms of whether they receive their preferred article choice. 3 out 5 experiments work with additional sub-randomization in which different sub-interventions are applied (psychological primes, cognitive load tasks, fact-checking).
Experimental Design Details
All interventions are conducted as part of online surveys with different online panel providers. In total, we conduct five experiments:

Experiment 1: Priming experiment. We randomly assign psychological primes to all study subjects. Psychological primes relate to forming ten sentences, each sentence contains of five words. There are two treatment groups (T1 and T2). In T1 all subjects receive words and sentences with non-religious content. In T2 subjects receive five sentences with religious and five sentences with non-religious content.
After the priming, study subjects select their preferred information sources for different topics. There are four types of sources (media, government, academic, religious) and four types of topics (2 health topics, 1 environmental topic, 1 religious’ topic). The order in which topics appear and in which the sources are listed is randomized.

Experiment 2: Information acquisition. Experiment 2 consists of two parts. Both parts involve a two step-process. In the first step, participants can select one out of four article titles that they would like to read. Titles reflect two different sources (Muhammadiyah vs. Nadul Ulama – two religious bodies in Indonesia) and two different opinions (pro and against smoking). In step 2 subjects are randomized into whether they receive an article that contains their preferred title (one out of four title options) or not (three out of four title options). The body of the article (three paragraphs) is identical for each subject and reflects arguments pro and against smoking. This experimental set-up is randomly assigned to 50 percent of all participants and we refer to this as PART A. For the other 50 percent of participants, the procedure is the same as in PART A, but with the exception that we indicate at both stages (article selection and article reading) whether an article was fact-checked for its content or not. At stage 1 about one fourth of the article titles are marked as fact checked; at stage 2 about 50 percent of the articles are marked as fact checked.

Experiment 3: Cognitive load. We experimentally alter the cognitive load of subjects. Cognitive load tasks relate to the memorization of specific numbers. There are two treatment groups (T1 and T2). In T1 study subjects are asked to memorize a 1-digit number, while in T2 study subjects are asked to memorize a 6-digit number. Following this, study subjects select their preferred article. Subjects can select from four articles titles with titles reflecting 2 opinions (pro or against smoking) and 2 sources (Muhammadiyah vs. Nadul Ulama). In a second step, subjects are randomized to receive their preferred article (one quarter of the title options) vs. a non-preferred article (three-quarter of the title options).

Experiment 4: Opinion vs. Source. Experiment 4 shall shed light on the role of the information source (religious source) vis-à-vis the informational direction (pro vs. against smoking) for the preference for information and andand information processing. Experiment 4 has three treatment groups (T1, T2, T3). In T1, study subjects see only the titles of articles though no source is displayed. Once subjects select a title, subjects are randomized into receiving their preferred article or not. The article they receive does not display any source. T2 is identical to T1 with the exception that once subjects are assigned an article to read they also see the source of the article. T3 is identical to T2 with the exception that the source of the article is already revealed at the initial stage when subjects are asked to select their preferred article.

Experiment 5: Article styles and information processing. In experiments 1 to 4 all subjects receive the identical article. Each article contains three paragraphs reflecting different views: One paragraph listing argument against smoking, one paragraph listing arguments why smoking might be permitted under certain circumstances, and one paragraph listing arguments why smoking should be allowed. In experiment 5, the articles displayed are changed such that they only present one position. This means that articles with titles indicating “pro-smoking” only list arguments in favor of smoking arguments, while titles indicating “against-smoking” only list arguments against smoking. Apart from the information (article) displayed, the experimental set-up of experiment 5 is identical to experiment 2 (part A).
Randomization Method
Randomization is done by the algorithm integrated in the online survey platform (UNIPARK)
Randomization Unit
Individual level
Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters
Sample size: planned number of observations
14500 individuals (respondents of online survey)
Sample size (or number of clusters) by treatment arms
Experiment 1: 1,500 individuals per treatment arm

Experiment 2:
o Part A: 1,000 individuals per treatment arm
o Part B: 1,000 individuals per treatment arm

Experiment 3: 1,500 individuals per treatment arm
Experiment 4: 1,000 individuals per treatment arm
Experiment 5: 750 individuals per treatment arm
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Minimum detectable effects were calculated using Stata’s power twomeans command. Given the absence of reference data it was assumed that the baseline mean in the control group is 0.5 for binary outcomes. MDEs are shown for binary outcomes only and relate to unconditional effects (without any covariates). MDEs are calculated for each experiment separately and pooled. • Experiment 1: 10.2pp • Experiment 2: 12.5pp (separately for part A and part B) • Experiment 3: 10.2pp • Experiment 4: 12.5pp • Experiment 5: 14.5pp • Pooled experiments (experiments 1 to 5): 4.6pp

Institutional Review Boards (IRBs)

IRB Name
German Institute for Global and Area Studies
IRB Approval Date
IRB Approval Number
Analysis Plan

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

Request Information


Post Trial Information

Study Withdrawal

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

Request Information


Is the intervention completed?
Data Collection Complete
Data Publication

Data Publication

Is public data available?

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials