Article Critique: Differentiation of Self, Stress and Emotional Support as
Predictors of Psychological Stress Lynne Gaynor EDF 5481 Florida International University

This critique discusses the article written by Romana C. Krycak, Nancy L. Murdock and Jacob M. Marszalek (2012) regarding the relations among stress, emotional support, and differentiation of self from the family of origin as predictors of psychological distress. The review begins with a summary of literature review; continues with the methods utilized, results, discussion and conclusion; and concludes with personal contributions.
Summary and Critique of Literature Review The topic discussed in the article, Differentiation of Self, Stress and Emotional Support as Predictors of Psychological Stress, focuses on both the measurement of stressful events and perceived stress and differentiation of self. The article discusses in length the relationship between differentiation of self, stress, social support, symptoms and psychological distress. According to Krycak, R.C., Marszalek, J. M. and Murdock, M. L., (2012) differentiation of self is defined as a person’s ability to maintain individuality while still being connected on an emotional level to his or her familial group. The researchers felt this topic was especially relevant among college students. Students who are higher in differentiation of self experience less psychological distress than students lower in differentiation of self. Thus, students who are higher in differentiation can succeed and achieve at their maximum potential. Researchers have also found the opposite to be true as well. Students who have a low differentiation of self experience more psychological distress and, thus, suffer negative consequences.
Krycak, R.C., Marszalek, J. M. and Murdock, M. L., (2012) reported in previous studies people at any point on the scale (of differentiation), if stressed sufficiently, could develop physical, emotional, or social symptoms. The higher the level of differentiation, however, the more stress required to trigger a symptom. Researchers tested this proposition and found two different interpretations of the relationship among stress, differentiation of self and symptoms. Some of the studies found that differentiation of self changes the relationship between stress and distress, whereas others show that differentiation of self mediates (or explains the mechanism by which stress affects) the relationship.
In another study researchers tested both the moderation and mediation hypotheses, and found that higher levels of differentiation were negatively related to both stress and psychological distress, but found that differentiation partially mediated the relationship between stress and adjustment. Krycak, R.C., Marszalek, J. M. and Murdock, M. L., (2012) reported that one line of research has focused on testing the nature of the relationships among differentiation of self, stress, and psychological distress, because Kerr and Bowen argued that stress would be experienced in the context of differentiation of self. The second line of research focused on testing other variables that should relate to differentiation of self as a way of further elaborating Bowen theory.
Based on Bowen’s theory and in previous research, researchers hypothesized that emotional support would mediate the relationship between differentiation of self and psychological distress. This indicates that one way in which differentiation influences psychological adjustment is related to how individuals of different levels of differentiation manage relationships. According to Krycak, R.C., Marszalek, J. M. and Murdock, M. L., (2012) researchers were also interested in the role that stress would play in this relationship; the researchers used path analysis to test the relationships among differentiation, stress, emotional support, and psychological distress. The studies that show that differentiation of self mediates the relationship were somewhat difficult to understand as compared to the studies that found that differentiation of self changes the relationship between stress and distress.
Summary and Critique of Methods The researchers in this study used a quantitative approach. Based on the relationship among the independent and dependent variables in the study the quantitative approach is appropriate and justified. The dependent variable being examined is psychological distress and the independent variables are stress, emotional support, and differentiation of self from the family of origin. The target population consisted of a sample of 200 college students. The authors did not introduce the sampling technique but based on my observations, they used convenience sampling. The problem with this sample selection is that it was not representative of the whole population and, thus, created a threat to external validity. Also, the students were not randomly selected. If the sample is not random the results will be non-generalizable.
The procedure of the study consisted of emails sent to professors, department chairs, and administrative assistants to distribute to their respective classes and list-serves. Participants were provided with a Survey Monkey link containing an electronic version of the surveys in the following order: DSI-R, CSI, HSCL-21, PSS, and the CBESS. This way of recruiting participants is probably problematic. A limitation to internet surveys is that samples are restricted to those with access to the technology and who choose to respond. According to our textbook, the multitude of potential responses in a web survey does not overcome the problem of sampling error due to lack of representativeness.
Measures were administered in the order specified in an attempt to ensure the quality of responses to longer measures such as the DSI-R and CSI. The study was anonymous, and the researchers used assigned random ID numbers; no identifying information was recorded. Included with survey measures was a statement detailing the voluntary nature of the study, potential risks and rewards for participation, and contact information for the university’s counseling center. Participants’ informed consent was indicated through completion and return of the survey measures. Participants were offered the opportunity to enter a raffle for one of five $10 gift cards to a local shopping center via a link to a separate online survey. It is possible that the motivation of the participants might be winning the gift cards and the participants may not particularly care about the subject matter of the study. Furthermore, they may provide inaccurate information at random. The data collection consisted of questionnaires. The Differentiation of Self Inventory-Revised is a 46-item self-report questionnaire. The disadvantage of this type of questionnaire is that it is online and inaccurate based on the fact that it is subjective and self-report. Furthermore, this long questionnaire may lead to participant fatigue, which would also affect the accuracy of their information. Items were rated on a six-point scale, where 1 means “not at all true of me” and 6 means “very true of me.” The DSI-R is comprised of four subscales: emotional reactivity (ER), I position (IP), emotional cutoff (EC), and fusion with others (FO). Internal consistency was Chronbach’s alpha and reported to be good. The Communication Based Emotional Support Scale (CBBES) is a 13-item self-report questionnaire that measures perceived emotional support, and is composed of items adapted from a previous questionnaire. Participants rate each item about their romantic partner on a scale from 1 to 5, where 1 means “almost never true” and 5 represents “almost always true.” Higher scores indicate more emotional support. In this study, items were adapted to refer specifically to the participant’s family support. Internal consistency reliability for this sample was also excellent. The Perceived Stress Scale (PSS) is a 14-item self-report questionnaire. Participants rated how often they experienced each item—such as “In the last month, how often have you felt confident about your ability to handle your personal problems”—on a five-point scale, where 0 means “never” and 4 means “very often.” Some items were reverse scored. High scores indicate higher levels of perceived stress.
Initial estimates of internal consistency reliability were good, with values ranging from .84 to .86. The College Stress Inventory (CSI) is a 21-item, self-report questionnaire that measures academic, financial, and social stressors of college students. An example of an item that measures financial stress is “difficulty paying tuition fees for the next term,” and an example of an item that measures social strain is “difficulty living in the local community.” The CSI had three subscales: academic stress, social stress, and financial stress. Participants respond to each item on a five-point scale, where 0 represents “never” and 4 “always.” Higher scores indicated more stress.
The Hopkins Symptom Checklist-21 is a short-form of the Hopkins Symptom Checklist. Participants rate 21 items, such as “feeling lonely,” on a scale ranging from 1 to 4, with the endpoints of “not at all” (1) and “extremely” (4). Higher scores indicate higher level of distress. The total score of the HSCL was used in the study, and the internal consistency coefficient was .91. According to Krycak, R.C., Marszalek, J. M. and Murdock, M. L., (2012), Vogel and Wei (2005) also reported an excellent full-scale internal consistency of .90. Lastly, Deane et al. (1992) tested the concurrent validity of HSCL-21 scores by assessing their correlations to similar measures, such as scores on the Brief Hopkins Psychiatric Rating Scale (BHPRS). They found moderate correlations, indicating moderate concurrent validity.
The measures used in the study were valid. First, the researchers examined construct validity by comparing a clinical sample to a non-clinical sample, and found that the non-clinical sample scored significantly higher than the non-clinical sample. Second, they compared scores on the HSCL-21 at entry into therapy and again at follow up. They found that scores on the HSCL-21 decreased significantly, which suggests a decline in psychological distress over time with therapy.
Summary and Critique of Results
Data screening—researchers conducted t tests to examine whether there were gender differences in the amount of differentiation of self, emotional support, perceived stress, actual stress, and psychological distress. According to Krycak, R.C., Marszalek, J. M. and Murdock, M. L., (2012) the research results indicated that the assumption of homogeneity of variance was met for differentiation of self, perceived stress, actual stress, and psychological distress, but not for emotional support. Taking this into consideration, t test results suggest that there were no differences between males and females on these constructs.
The results of the study confirmed a number of knowledgeable findings. The researchers demonstrated that differentiation of self, emotional support, and measures of perceived stress and stressful events were related to psychological symptoms in theoretically predicted ways. According to Krycak, R.C., Marszalek, J. M. and Murdock, M. L., (2012), path analytic findings demonstrated that actual stress events were directly related to symptoms, but also indirectly related, through perceptions of stress. All of the tables used to display statistics and correlations were clearly labeled, however, the diagram of the hypothesized path model was a bit confusing to interpret. The researchers provided sufficient statistical information and collected all of the data needed to measure the variables in question. Summary and Critique of Discussion/Conclusion To test whether or not the conflicting findings in the literature resulted from a difference in what was being measured as ‘‘stress’’ (i.e., actual vs. perceived), the analyses of Murdock and Gore (2004) and Skowron et al. (2004) were replicated with the current sample using the same measures of stress used in the original studies (PSS and CSI, respectively). Researchers did not find support for Murdock and Gore’s hypothesis that differentiation of self moderates the relationship between perceived stress and psychological distress; however they did find partial support for Skowron et al.’s hypothesis that differentiation of self mediates the relationship between stress events and psychological distress. At the same time, path analytic results indicated that differentiation mediated the relationship between stressful events and perceived stress, suggesting that the type of stress measure used does matter.
The conclusions drawn by the researchers seemed reasonable and I agree with the researchers conclusions that emotional support may be a protective factor against the development of psychological difficulties for a client, but more research is needed to replicate these findings. According to Krycak, R.C., Marszalek, J. M. and Murdock, M. L., (2012) a limitation of this study was that a volunteer sample of college undergraduate and graduate students was used. Additionally, the sample consisted largely of Caucasian females. Given these factors, the results of this study may have been sample-dependent and external validity of the results may be somewhat limited. Also, according to Krycak, R.C., Marszalek, J. M. and Murdock, M. L., (2012), demographic information collected about the sample did not include type of family of origin structure (blended families, intact families, etc.), or whether or not participants were receiving therapy to cope with stress. Knowing more about the characteristics of the sample may have aided in interpretation of results, particularly with regard to other factors that may protect against the negative effects of stress.
Personal Contribution
After reviewing the article, I felt it was extremely thorough. The findings of the article did directly relate to my area of interest. I feel through my own personal experiences, trials and tribulations; differentiation of self played a substantial role in my psychological well being as a young woman and college student. The research topic of differentiation of self as it relates to stress and distress is relevant to my future career path, as I will be counseling young men and women, school-aged and young adults. As a counselor, my clinical role may be working with clients to help increase their level of differentiation of self. As the therapeutic process moves forward hopefully helping clients increase their level of differentiation can result in the client’s ability to better form and maintain a good support network. Which will ultimately help the client in times of distress.
In reading the article I learned that there is a direct correlation between differentiation of self, emotional support, and measures of perceived stress and stressful events related to psychological symptoms. Bowen’s family systems theory is based on the premises that the individual’s primary purpose is to differentiate him or herself from the family of origin and that previous generations’ behavior and level of functioning are associated with this process in the current and future generations. Theoretically, differentiation of self pertains to the ability to maintain emotional objectivity during high levels of anxiety in a system while concurrently relating to key people in the system (i.e. family members, peers, coaches, professors, etc).
There were several limitations in the article. First, the students were not randomly selected. If the sample is not random the results will be non-generalizable. I would suggest selecting students at random which would allow for representation of the entire population. Next, I would suggest using surveys that are not internet based, as the online questionnaires are restricted to those with access to the technology and who choose to respond. I would also suggest shortening the questionnaires. Long questionnaires may lead to participant fatigue, which will also affect the accuracy of their information.
Third, the sample consisted largely of Caucasian females. Perhaps researchers could have broadened their sample population in terms of gender and ethnicity, which in turn, would have been a more realistic representation of the entire population, rather than limited to strictly to Caucasian females. And, finally, the demographic information collected about the sample did not include type of family of origin structure or whether or not participants were receiving therapy to cope with stress. Perhaps, knowing more about the characteristics of the sample may have helped in the interpretation of the results, particularly with regard to other factors that may protect against the negative effects of stress.

References
Krycak, R.C., Marszalek, J. M. and Murdock, M. L., (2012). Differentiation of Self,
Stress and Emotional Support as Predictors of Psychological Stress. Journal of Contemporary Family Therapy, 34, 495-515.