A Department of Homeland Security Center of Excellence led by the University of Maryland

Homegrown Radicalization and the Role of Social Networks and Social Inclusiveness in the United States


Homegrown Radicalization and the Role of Social Networks and Social Inclusiveness in the United States

Investigators: 
Other START Researchers: 
Clark McCauley

Project Details

Abstract: 

This project involves a telephone survey of a sample of approximately 1,000 Americans, using a survey instrument followed by an online panel survey. Variables, for example, sought to gauge a respondent's perceived level of political and economic deprivation within their community; their level of religiosity, political efficacy; time in America (citizen/ non-resident; 1st/2nd generation etc.) The telephone survey also involved an over-sample in areas of high population density in order to gauge the role played by social inclusiveness within the radicalization process. Social network analyses measuring the level of inclusiveness of networks were conducted, as well as identifying the "nodes" of these social networks.

Primary Findings: 

Muslim-American respondents were more likely to indicate that they generally approve of the use of violence by a group to achieve political aims than those among the general population. However, those among the Muslim oversample were also the most likely to report strongly disapproving of the use of violence to achieve political aims. 

Muslim respondents were nearly twice as likely to report that more than ten of their friends lived in their neighborhood. Conversely, respondents within the general population were more likely to report that none of their friends lived in their neighborhood.  Muslim respondents and subjects within the regional oversample were more than twice as likely as those in the general population to indicate that people who share their ethnic background give them a sense of community. They were also significantly more likely than the general population to indicate that they get a feeling of belonging or sense of community from their place of worship, people online, and people in their neighborhood. 

Among the general population, younger respondents were more likely to report that people who share their ethnicity provided them with the sense of community. Across all age cohorts, Muslim respondents were far more likely than those in the general sample to indicate that people who share their ethnic background provide them with a sense of community. 

A majority of respondents in all three populations reported disagreeing that they'd prefer to live in the U.S. under religious law rather than secular law. In fact, those within the general population were marginally more likely than those in the Muslim and regional oversamples to report agreeing that they'd prefer to live under religious law.

Methodology: 

These data are a result of a national telephone survey written and administered by the Public Opinion Research Laboratory at the University of North Florida in Jacksonville. Using Random Digit Dial (RDD) technology, approximately 100,000 phone numbers were dialed, yielding approximately 1,622 total completed interviews. In addition, 173 self-reported Muslim households were oversampled: 94 nationally and 79 specifically targeted in four metropolitan areas around the country: Detroit, MI, Jacksonville, FL, Brooklyn, NY, and Atlanta, GA. These districts were chosen as a result of their comparatively high concentration of mosques per capita in an effort to determine if attitudes, behaviors, and perceptions of social inclusion differ in areas densely populated with a more active Islamic community. Given the target populations, the survey instrument was translated into Spanish, Urdu, Hindi, Farsi, and Arabic by native speakers of each respective language. Respondents who indicated an inability to complete the survey in English were transferred to trained interviewers capable of completing the survey in one of the five aforementioned languages. The final data includes information for three distinct populations: 1) General Population (N = 1,622): 1,579 Non-Muslims and 43 Muslim-Americans sampled randomly across the U.S. with strata quotas proportional to Census division; 2) Muslim Oversample (N = 137): 43 Muslim-Americans from General Population sample plus an additional 94 Muslims oversampled randomly across the U.S. with oversample strata quotas proportional to Census division; 3) Regional Oversample (N = 79): 79 Muslim-Americans oversampled within 4 metropolitan districts across the U.S. (Detroit, MI, Atlanta, GA, Brooklyn, NY, and Jacksonville, FL) with high concentrations of mosques per capita. The data was analyzed using SPSS/PASW analytics software.  

Timeframe

Project Period: 
June 2008 to October 2011