Unmatched Count Technique
Also known as: Item Count Technique, List Experiment, Unmatched Block Design
The unmatched count technique (also called the item count technique or list experiment) is an indirect survey method for estimating the prevalence of sensitive attitudes or behaviors while protecting respondents' privacy. Respondents are randomly assigned to one of two versions of a question. The control group sees a list of several non-sensitive items and reports only how many of them apply to them; the treatment group sees the same list plus one additional sensitive item and likewise reports only the count. Because respondents report a number rather than which items apply, no one's answer reveals their stance on the sensitive item. The estimated prevalence of the sensitive attribute is simply the difference in mean counts between the treatment and control groups. By breaking the link between an individual and the sensitive item, the technique reduces social-desirability bias for topics like prejudice, illegal behavior, or stigmatized attitudes, as documented in validation work by Holbrook and Krosnick.
Key highlights
- Protects respondent privacy, reducing social-desirability bias.
- Yields population prevalence estimates via a simple difference in means.
- Applicable to many sensitive attitudes and behaviors.
- Extensible with regression and multivariate estimators.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Use the unmatched count technique when you need to estimate the prevalence of a sensitive attitude or behavior -- prejudice, stigmatized opinions, illegal or embarrassing conduct, turnout misreporting -- and direct questions are likely to be distorted by social desirability. It provides population-level estimates with strong privacy protection. It is less appropriate when individual-level measurement is required (it yields aggregate prevalence), when sample sizes are too small for the difference-in-means to be precise, or when control items can be poorly chosen (risking ceiling and floor effects that compromise anonymity and validity); Holbrook and Krosnick's work also shows it does not always reduce bias. Careful list construction, randomization, adequate sample size, and statistical checks for design effects are essential.
Strengths & limitations
- Protects respondent privacy, reducing social-desirability bias.
- Yields population prevalence estimates via a simple difference in means.
- Applicable to many sensitive attitudes and behaviors.
- Extensible with regression and multivariate estimators.
- Estimates prevalence at the aggregate level, not individual status.
- Requires larger samples than direct questions for the same precision.
- Validity depends on careful control-item selection to avoid ceiling/floor effects.
- Does not always reduce bias, as validation studies have shown.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
How does the list experiment protect privacy?
Respondents report only how many items on a list apply to them, not which ones. Half receive a list that secretly includes a sensitive item and half receive the same list without it. Because the answer is a single count, no individual's stance on the sensitive item can be inferred, removing the incentive to misreport that direct questions create.
How is prevalence estimated?
The prevalence of the sensitive attribute is the difference between the mean count in the group whose list included the sensitive item and the mean count in the group whose list did not. Because the groups are randomly assigned and otherwise identical, this difference-in-means recovers the share of people who endorse the sensitive item without identifying any individual.
What can compromise the technique?
Poorly chosen control items can create ceiling or floor effects -- if someone reports all or none of the items, their sensitive status is revealed, breaking anonymity. The method also needs larger samples than direct questions for the same precision, and validation work shows it does not always reduce bias, so careful list design and benchmarking are essential.
Sources
- 1.Holbrook, A. L., & Krosnick, J. A. (2010). Social desirability bias in voter turnout reports: Tests using the item count technique. Public Opinion Quarterly, 74(1), 37-67.
- 2.Ross, L., Greene, D., & House, P. (1977). The 'false consensus effect': An egocentric bias in social perception and attribution processes. Journal of Experimental Social Psychology, 13(3), 279-301.
You have read it. What now?
Cite this page
ScholarGate. (2026, June 23). Unmatched Count Technique. ScholarGate. https://scholargate.app/social-psychology/unmatched-count-technique