Threats to Internal Validity

The aim of my research study is to assess the self-perceptions adolescent girls have of how their single family homes affect their school performances. Data will be gathered via interviews with interviewers asking a set list of questions. Like any study, there is a risk for the internal validity of the research to be weakened if extraneous and confounding variables impact the dependent variable. To take all precautions to combat this, these threats are being examined in effort to control the extraneous and confounding variables.

Depending on how recent the adolescent female has found herself in a single family home may have an affect on her how she perceives her school performance. If a girl is adjusting to the new and different home environment (whether it be caused by a recent death, divorce, or other happening), she may self-report that the single parent home is having more of an influence on her school performance than if she had completely adapted to the single parent setting. This type of threat would be categorized has history since time is associated with the independent variable.

Although I would do my best as a researcher to ensure adolescents were selected from schools with similar socioeconomic statuses, school contexts (similar school faculty and staff quality and student populations), and geographic settings, selection still poses a threat to the internal validity of the study. I believe that though measures will be taken to minimize the validity of this study, in a real world setting, no two places are the same, so the effect of the selection settings will inevitably skew results.

My assumption is that some of the interviews with participants may be long, and for the early adolescents (12-14 year olds), this time may effect their answers over the course of the interview session. This type of threat is called maturation. In addition, the cross-sectional design of research will evaluate the self-perceptions of all adolescent girls ranging from ages 12-18 year old. Obviously, the older the adolescent, the more socially, emotionally, mentally, and physically she has developed. If not controlled, this will weaken the internal validity of the research.

When evaluating how instrumentation may influence the internal validity, a couple of threats may be involved. First, whenever one is self-reporting, the standard is completely subjective meaning questions are interpreted in the context of the person who is answering them. In addition, the questions that are asked may not allow for the teenager to expound fully on how she perceives her single parent home affecting her school performance. Questions can be limiting to the extent of the self-perception.

To execute high quality research, I plan to have a team of experimenters assisting with the interviews. Though there will be a list of set questions, the manner in which they are asked depends on the interviewer. This effect, called Experimenter effect, even involves some characteristics of the interviewer that cannot be prevented (age, sex, race, physical appearance). Some of the qualities of the interviewer that can be prevented (behavior, conduct) can be taught during a training session with all interviewers.

As an ethical research practitioner, I will not force a girl to be in the study if she is not a willing participant. I understand that the topic of home life can be a sensitive and vulnerable topic for some adolescent girls to speak about, so the ones who participate may be ones who feel their performance in school is less affected by their single parent living situations. Furthermore, even if some girls who are interviewed do perceive their single parent family having an effect on their school performance, particularly a negative effect, they may not be likely to share this information in fear of shame and embarrassment. I think this proves that a trusting relationship must be built between interviewer and participant to ensure accurate information is gathered.