Terrorism Suspect Religious Identity and Public Support for Controversial Detention and Interrogation Practices

Download data and study materials from OSF

Principal investigator:

James Piazza

The Pennsylvania State University

Email: jpiazza@psu.edu

Homepage: https://sites.google.com/site/jamesapiazzapennstate/

Sample size: 1135

Field period: 8/30/2011-1/3/2012


This study proposes to examine the effects that the religious identity of a terrorism suspect has on public support for the application of controversial interrogation techniques and detention practices that have become part of United States counterterrorism policy since the September 11th 2001 terrorist attacks. It tests the hypothesis that the public is more permissive of physically abusive interrogation of Muslim terrorist suspects and is more accepting of indefinite detention and transfer of accused terrorists to military commissions for suspects identified as Muslims or as associated with extremist Islamic movements. This study executes an original national survey and finds that respondents are generally more supportive of subjecting terror suspects with stereotypical Muslim names or that are associated with a radical Islamic terrorist group to harsher treatment than non-Muslim suspects associated with domestic, right-wing terrorist movements. This effect is found to be strong for extraordinary detention practices such as indefinite detention and denying suspects access to legal counsel and civilian criminal courts. These results are robust to the inclusion of controls for respondent age, income, education, race, gender, attitudes towards race, personality type, religion and employment situation.


H1: Respondents are more likely to support harsh detention and interrogation tactics when directed against Muslim terrorism suspects.

H2: Respondents are less likely to support harsh detention and interrogation tactics when directed against unidentified or right-wing terrorism suspects.

Experimental Manipulations

The proposed study executes a survey experiment involving four treatment vignettes and one control vignette and 17 survey questions administered to 1,050 respondents. Respondents are randomly assigned to one of the five treatments which depict a short AP newswire blurb describing an arrest of two terrorist suspects in suburban Chicago. The treatments are identical to one another except they vary the names of the suspects (stereotypical Arabic/Muslim vs. Anglo-American) and the names of the terrorist movement the suspects are alleged to be members of (radical Islamists vs. right-wing American extremist). The control vignette omits any identification of the suspect names or groups. All respondents are then asked 13 questions rating their support for / approval of controversial interrogation and detention practices (10 interrogation practices, including the use of physical abuse of suspects, and three detention practices, including indefinite detention of suspects) that have been used by U.S. counterterrorism officials since the 9/11 terrorist attacks. The study employs a series of ordered logistical regression estimations to determine the effect of treatment on support for interrogation and detention practices.


The main independent variables of the study are dummy indicators for the four treatment vignettes and the one control. The dependent variables are respondent levels of support for 13 controversial interrogation and detention practices measured on a five point Likert scale. Included in some estimation models are control variables for respondent household income, employment situation, age, gender, political ideology, religion, residence, marital status and parental status.

Summary of Results

The results indicate strong in-group and out-group effects consistent with Integrated Threat Theory, at least for respondent tolerance for extreme detention of suspects. Respondents in the “Muslim” treatment groups – those that read news stories where either the suspect or the suspected terrorist group was identified as Muslim or radical Islamist – were significantly more likely to condone extreme detention practices against the suspects than those exposed to the other treatments and the control. Moreover, respondents in the treatment group where the suspected terrorist movement was identified as a right-wing domestic American organization were significantly less likely to support extreme detention. No significant relationships were found between treatments and support for extreme interrogation of suspects. Results were robust to the inclusion of all control variables and some alternate model specifications.


Piazza, James A.  2014.  "Terrorist Suspect Religious Identity and Public Support for Harsh Interrogation and Detention Practices."  Political Psychology.  Forthcoming.