US tab

An Analysis of the Ecological Theory of Research
Participation Applied to a Sample of Young
Adult Males

By: Lisa M. Soler, RN and Adriana Ramirez | Mentor: Dr. Michael J. Rovito

Background

Problem Statement

Longitudinal studies are of significant value in interventional, evaluational, and developmental studies, and are also one of the stronger sources of evidence for suggesting causal relationships (Marcellus, 2004). However, these studies are costly and are particularly vulnerable to attrition. The systematic introduction of bias and error (specifically non-response and refusal bias) into the study poses challenges to its internal, external, and statistical validity (comprehensive validity) (Marcellus, 2004). Such challenges to comprehensive validity arise because of character differences between participants who drop out of the study and those who do not, which in turn leads to sample bias through the misrepresentation of the targeted population (Marcellus, 2004). Additionally, the possibility of a reduction in statistical weight arises when the sample size experiences significant loss (Marcellus, 2004). Unfortunately, limited attention is given to this topic despite the aforementioned hindrances on the quality of information obtained from longitudinal research.

Causes of Attrition

Ribisl et al. (1996) and Marcellus (2004) recognize participant, researcher, and contextual reasons as primary sources of attrition. In health promotion research, factors under the "participant-related" category (e.g., gender, ethnicity, etc.) are largely responsible for loss to follow up. Vanable et al. (2002) observed that participants who complete a study are typically among the older age range of the targeted demographic than those who do not complete the study. This is attributed to young individuals' self-perception of their assumed healthiness; therefore, younger people may perceive health concerns as of limited relevance to their lives.

Ethnicity has also been observed as a contributing factor to attrition. Dolinsky, Armstrong, and Ostbye (2012) found non-white/non-Hispanic children were more likely to drop out of a study than white non-Hispanic or Hispanic children. Hui et al. (2013) found similar patterns in their study with oncology patients, although Hispanics were shown to have a higher drop-out rate. Considerations from both studies suggest minority races in general exhibit higher attrition than whites. Reasons behind this remain largely unknown, but socioeconomic and/or cultural value variables may be related.

Additional participant related reasons for attrition include gender and school performance. Mein et al. (2012) examined higher attrition rates among males than females. Further, Fröjd et al. (2011) indicated that lower school performance strongly correlates with attrition, particularly in a study targeting an adolescent population for a mental health cohort. The author noted that a higher GPA resulted in a lower incidence of attrition.

"Researcher-related" attrition refers to a situation in which the primary investigator forms a role that is perceived as paternalistic or "too authoritative," which creates a barrier between researcher and participant(s) (Gross and Fogg, 2001). Focusing on the researcher's needs to obtain data, as opposed to the needs of the participant, can contribute to disengagement from the study. One must take caution not to elicit feelings of inferiority in partakers but instead to incorporate an element of active participation (Marcellus, 2004). Developing a "horizontal" relationship, in which participants feel equal to the researchers, eliminates participants' perceptions that they are test subjects while also increasing assurance in their significance to the study.

Finally, "contextual reasons" contribute to a study's attrition rate. According to Barry (2005), collecting data deemed sensitive in nature can result in participants' refusal to adhere to study requirements such as questionnaires or surveys. Participants may feel uncomfortable disclosing private information, particularly in regard to sexual health. This may lead to tailored answers or complete withdrawal from the study. Moreover, setbacks frequently emerge when face-to-face contact is limited. Methods that decrease participant-researcher interaction result in loss of interest and engagement. Booker et al. (2011) found that face-to-face interviews increase retention rates by 24 percent. The authors of this article suggest this may be because alternative locations and methods of data collection were offered. Therefore, one must consider how the target audience perceives the study itself: it must not be overly taxing for participants, while still carrying out the necessary agenda of the study

Other contextual reasons focus on time commitment and perceived effort for the participant. While it may be essential for some researchers to meet participants in person, studies that make such interactions too time consuming or demanding may cause participants to become irritated with the perceived rigors of the study. This problem may be amplified if subjects are required to use excessive cognitive ability or time when filling in data, such as in a lengthy questionnaire (Fröjd et al., 2011). If individuals in the study feel they are doing "too much work," one runs the risk of compromising data from their hasty responses. Additionally, it is imperative that all study materials are developed in such a way that no individual be excluded or at a disadvantage due to his/ her cognitive ability.

Observations >>