Qualitiaive Interviewing Plan -- Vignettes

The Census Bureau plans to conduct additional research under the generic clearance for questionnaire pretesting research (OMB number 0607-0725). The goal of this research is to use vignettes to explore complex survey concepts. The results will be used to demonstrate the effectiveness of including vignettes as part of survey questionnaire presenting methods.

Vignettes, which depict hypothetical situations, are a common tool for survey pretesting. They allow researchers to evaluate complex situations and concepts without the onus of having to recruit people in those particular situations. Vignettes also are useful when evaluating potentially sensitive topics. Respondents may not be willing to self-report certain sensitive or risky behaviors, but they may be more willing to talk about hypothetical situations depicting such behaviors. The current research is intended to explore how vignettes can be useful in evaluating concepts about sensitive topics. More specifically, we will identify and explore how teens classify their contacts with online strangers and the degree to which they are aware of the danger in such interactions.

This study grew out of using vignettes to pretest Internet Predation questions on the NCVS. Those questions, which will be asked of kids and teens between the ages of 12 and 17, will collect information on potentially risky online behaviors. Because of the logistical difficulties of recruiting kids and teens who had engaged in risky online behaviors, we used a series of hypothetical vignettes to explore some of the important concepts involved in such risky behaviors. Without the vignettes, we would not have been able to collect much information on how kids and teens think about their online activities that may result in the misreporting of target behaviors.

The current research, which will expand on the limited information we collected in the pretesting, will be a mixed-design qualitative and quantitative study. Participants will classify vignettes depicting online contacts with strangers and online contacts with non-strangers as either being appropriate and harmless or inappropriate and potentially dangerous. Participants will also answer open-ended questions about why they feel these contacts are or are not dangerous. In particular, we will be looking for specific themes in what determines the potential danger in an online interaction. We will then classify and analyze these themes. The results of the vignette classification task and the open-ended questions will shed light on the usefulness of vignettes at exploring and identifying how teens conceptualize their online relationships.

Between March and August 2009, staff from the Statistical Research Division (SRD) will conduct a maximum of 60 qualitative interviews with kids and teens between the ages of 12 and 17. To the greatest extent possible, we intend to gather a sample that is balanced across race, ethnicity, and gender.

We will depart slightly from our traditional means of recruiting participants for this study. Normally, we recruit a number of our research participants through ads posted in free local newspapers, ads posted on “online classifieds” websites (Craigslist), or through community organizations. Because we are investigating potentially risky contacts with online strangers, placing online ads and newspaper ads requires that parents respond to online strangers. This recruiting method could introduce a potentially problematic confound in our results. We would be recruiting participants whose family already engages in the risky behavior we intend to explore. As a result of this potential confound, we intend to recruit our participants through local schools, local community youth programs, and other local youth organizations (i.e., church youth groups).

We will conduct our interviews using a semi-structured qualitative interview protocol (enclosure 1). We will present participants with a series of 12 different vignettes, depicting online communications between a teen and someone else (enclosure 2). These vignettes will vary on two important dimensions. First, these scenarios will involve interaction with people the teen knows only online or interaction with people with whom the teen has some other offline connection (such as through friends or family). Second, the scenarios will either depict the teen as not responding to the online stranger (not complicit in the interaction), or depict the teen responding to the contact (complicit in the interaction). Pretesting results of the Internet predation questions suggested that these two dimensions were important to how kids and teens viewed online communications. We want to explore the degree to which these two dimensions affect the perceived danger in online communications. However, we are not limiting our results to these two themes. We hope to be able to capture how kids and teens weigh the various factors that contribute to the feeling that some online interactions could be dangerous.

We also intend to collect some basic demographic information (enclosure 3) that will be important to our results. More specifically, we will collect information on the type of school the kids and teens attend (public, private, or private and religiously affiliated), the frequency of online use, online activities (gaming, blogging, web surfing, etc.), having a personal webpage or social networking site page, where they have online access (at home in common area, at home in bedroom, at school, etc.), race/ethnicity and gender.

Interviews will be conducted in the Washington DC metropolitan area, either in the Census Bureau’s cognitive laboratory or locations convenient to the participants

Interviews also will be tape-recorded, with the parent/gaurdian permission and minor assent, to capture the responses we will be coding and analyzing. All participants will be informed that their response is voluntary and that the information they provide is confidential (see enclosure 3 and enclosure 4 for the parental consent and minor assent forms, respectively). Participants will receive $15 for their participation.

The estimated time for completion of each interview is one hour. Thus, the estimated burden for this research is 60 hours.

The contact person for questions regarding data collection and study design is:

Jennifer Beck

Center for Survey Methods Research

Statistical Research Division

U.S. Census Bureau

Room 5K112D

Washington, D.C. 20233

(301) 763-1736

jennifer.l.beck@census.gov