Writing About My Data: Describing My Study Methods

 Sample

The sample studied included adults over the age of 18 residing in the United States who responded to the online Outlook on Life survey in 2012. 2294 respondents completed the first wave of the study; of those, 1601 completed the second wave also. Respondents were categorized by ethnic groups including Blacks (55.7%), Whites (35.5%), Latinos (5.4%), Other (2.0%), and Mixed Race (1.4%). My sample included all respondents who answered the questions relevant to my analysis.

Procedure

Data were collected by GfK Knowledge Networks for the University of California, Irvine via online surveys in two waves: wave one was conducted on August 2012; those who completed wave one were then asked to complete wave two on December 2012. Individuals were initially contacted via phone or postal mail and asked to fill out the online survey.  The survey was conducted to study the influences of variables such as political and sexual orientation, religiosity, and other beliefs influence political and social attitudes. 

Measures

My explanatory variable was Religiosity, which was calculated as a composite of several questions:  “what is your religion?”, “are you born again?”, “do you approve of women clergy?”, “how often do you attend church?”, “are you an active member of your church?”, “do you approve of gay clergy?”, “do you consider homosexuality immoral?”, and “do you approve of laws protecting gays?”.  The composite was divided by the number of valid responses to account for non-answers, giving an index ranging from 0.0 to 2.0.

My response variable was Racial Acceptance, which was also a composite of 14 questions, including: “were you ever willing to date outside your race?”, “how do you feel about biracial children with [blacks/Latinos/whites/Asians]?”, “what is the predominant race of your church?” (1 point if it was different from respondent’s ethnicity), “rate [Native Americans/Latinos/whites/blacks/Asians] on a scale of 1 – 100”, “should blacks teach their children against racism?”, “should blacks segregate from whites?”.  This composite was also divided by the number of valid responses, giving an index ranging from 0.0 to 1.0. The two indices are quantitative variables, so a Pearson Correlation was run to determine possible relationship.