What is a List Experiment?
When asked about sensitive behavior, people may give the answer they think is expected rather than the truth. This is known as social desirability bias and can lead to under-reporting of sensitive behaviors, particularly where disclosure could bring stigma, fear, or other consequences.
The List Experiment, also known as the Item Count Technique, is an indirect questioning method designed to reduce this bias while protecting respondent privacy. Instead of answering yes or no, respondents hear a short list of statements and report only how many apply to them. The sensitive statement is mixed with neutral statements, so neither the researcher nor enumerator can tell whether it applies to an individual respondent.
At 60 Decibels, we have used List Experiments to assess potential human rights risks across rice, coffee, and cocoa supply chains in multiple geographies. Companies have used these findings to monitor the prevalence of risks across their supply chains, inform their due diligence efforts, and design targeted remediation.
How does a List Experiment work?
Respondents are randomly assigned to one of two groups:
Control Group
Receives a list of three non-sensitive, or innocuous, statements.
Treatment Group
Receives the same three statements, plus one sensitive statement.
Example:
Control List
- 1It is common for family members to help on the farm.
- 2It is common for men and women to perform similar types of tasks on the farm.
- 3It is common for farm laborers to decide their pay.
- 0
- 1
- 2
- 3
Treatment List
- 1It is common for family members to help on the farm.
- 2It is common for men and women to perform similar types of tasks on the farm.
- 3It is common for farm laborers to decide their pay.
- 4SensitiveIt is common for women to be paid differently than men for the same type of work.
- 0
- 1
- 2
- 3
- 4
Rather than asking which individual statements apply to them, we ask only how many statements apply. Because the researcher sees only the total count, they cannot tell whether any individual respondent selected the sensitive statement. This gives respondents privacy and can make it easier to answer honestly.
At the aggregate level, we can estimate the prevalence of the sensitive statement by comparing the average number of statements selected by the treatment group with the average selected by the control group. Because the sensitive statement is the only difference between the two lists, the gap between the averages estimates its prevalence.
This is particularly valuable where direct questioning leads respondents to under-report.
1.6statements
Control average
1.9statements
Treatment average
0.3difference
30% estimated prevalence of the sensitive behavior