Limitati[)ns and futur~ r~s~arch

First, it is worth noting that limitations in the generalizability of this study present stu may have resulted from sample selection. In the present study, subjects voluntari­ earlier in 1 ly participated in this research project; hence, they may be representative of a cant marit~

subpopulation that has certain unique traits such as greater life satisfaction or a responses 1 greater willingness to help others than the general population. A higher propor­ argued tha tion (43%) of subjects were recruited from churches; hence the results are not to the indi necessarily generalizable to the general population. In. addition, a higher propor­ markedly f tion (64%) of old subjects came from senior centers; therefore, they may not be behaviors proportionally representative of the old population. Further, research participants across the in the present study were predominantly middle class, in their first marriage, in ful events good physical health, and Protestant; hence they may not be broadly representa­ Launier, 1~

tive of the general population. product of

A brief evaluation of the limitations of the measures used in the present s­ to discuss.

tudy should be included. For the assessment of marital coping behaviors, research researchers participants were asked to recall the most serious recurring marital conflict they same situal have experienced, then, decide to what extent they used the items listed in re­ ences from sponding to that stressful event. Readers may wonder whether participants would iors for acl follow the instructions for responding. If participants did not follow the instruc­ tal coping tions or could not recall a significant recurring marital conflict, responses to items ated by re:

would reflect a generalized marital interaction style rather than a specific coping conflicts wi response to marital conflict. In a similar vein, appraised stressfulness and ap­ ized scenal praised controllability of marital conflict situation may reflect one's general evalua­ results froll tion of marriage and general locus of control. In addition, as mentioned above, generated 1 both appraised controllability and appraised stressfulness were assessed with a sin­ Futur gle item; as a result, the validity of these two measures are questionable. To marital COl remedy the dubious psychometrics of cognitive appraisal measures, efforts need to can corrob

be made to design a larger scale with multiple items. search will

A primary focus in the present study involved the assessment of marital cop­ a more COll ing behaviors. Therefore, how to precisely assess marital coping behaviors is an Future res important issue for future research. In conducting research on coping with marital of cognitive conflict, one choice is to ask respondents to generate conflict from their own mar­ social supp riage experiences and respond to the extent they deal with this conflict as the tures of the

of this study ects voluntari­

ientative of a isfaction or a igher propor­

suIts are not igher propor­

may not be

I participants marriage, in

I representa­

e present s­

leS, research

~onflict they isted in re­

pants would the instruc­

,es to items eific coping ss and ap­

:ral evalua­

ned above, with a sin­

lable. To ts need to

Hital cop­

iors is an th marital own mar­

ct as the

OOA '

present study conducted. This strategy has methodological limitations as discussed earlier in that respondents may not follow the instructions or can not recall signifi­

cant marital conflict events. Another problem is that if a study is based solely on responses to questions about conflicts generated by respondents, it can readily be argued that the respondent's responses are idiosyncratic, reflecting their reactions to the individual conflicts they choose to discuss. Since the conflicts would differ markedly from one another, it would be premature to conclude that marital coping behaviors reflect general tendencies which might be related to other variables across the respondents. Coping behaviors may be a function of the type of stress­

ful events (e.g., loss, threat, or challenge) which people encountered (Lazarus &

Launier, 1978; McCrae, 1982). Differences in marital coping behaviors could be a product of the different types of marital conflict situations that respondents chose to discuss. To control for systematic differences in the types of conflict situations, researchers can provide standard scenarios so that respondents all react to the same situations. Using this technique, researchers are able to draw valid infer­

ences from their data, but they are not informed about respondents' coping behav­

iors for actual problems. Therefore, future research on this issue can obtain mari­

tal coping data by using both standard marital conflicts and marital conflicts gener­

ated by respondents from their own experiences. Information about actual marital conflicts within marriage can be solicited, and information obtained from standard­

ized scenarios used to ensure that valid inferences are drawn. In particular, the results from the standard scenarios can be used as a check on the marital conflicts generated by the respondents.

Future studies can benefit from the use of other kinds of measures to assess marital coping behaviors besides self-report. Observational research, for example, can corroborate the information obtained from self-report measures. Future re­

search will also do well to obtain data from significant others or spouses to obtain a more complete picture of the influences of variables on marital coping behaviors.

Future research in this area should focus on the possible multidimensional nature of cognitive appraisal and social support. For example, as Thoits (1982) showed, social support can be characterized in terms of types, amounts, sources, and struc­

tures of the support networks; as Lazarus and Folkman (1984) suggested, the indi­

clinically applicable typology.

American Journal

Family Therapy,

Consul