Research Article Panel Presentation
Marital/ Couples Assessment Tests
Freeston, M.H., & Plechaty, M. (1997). Reconsideration of the Locke-Wallace Marital Adjustment Test: Is it still Relevant for the 1990S? Psychological Report, (81) 419-434.
Family therapist in clinical settings extensively uses marital Assessments. They are important for finding how compatible and satisfied married couples are in their relationship. The Locke-Wallace Marital Adjustment test focuses on aspects of marriage such as general marital satisfaction or quality deciding factors on life issues and relationship style.
The Locke-Wallace Marital Adjustment Test has been widely used in the marital quality over the last 30 years and it continues to be used despite the development of other scales (Freeston& Plechaty, 1997). The test The Locke-Wallace test consists of 15 items and is rapid to admire and score despite the unequal weights for different items. Rapid assessment instruments are needed in many clinical settings to measure important variables related to research hypothesis or a control. In the past the test has been criticized on a number of issues such as the scale content and the underlying construct of the test and the psychometric issues. Because there has been few empirical examinations of these criticisms the article intend to contribute empirical data in support of the continuing use of the Locke-Wallace Marital Adjustment Test.
The Locke-Wallace originally indicated high internal reliability (.90) but no test retest stability. Later criticism of the test suggests that reliability is weak base on this estimate of Cronbach alpha of .77 and variability in the item total correlations (Freeston & Plechaty, 1997).
The Locke-Wallace Scale has not been widely factor analyzed and no subscales have been proposed in contrast with the Dyadic Adjustment Scale, which has been repeatedly analyzed although the subscales and have been difficult to replicate. (Freeston & Plechaty, 1997). The Separate analyses of responses from men and women have yielded different structures, which make it difficult to understand the underlying structure.
The study examined the reliability, criterion related validity, factor structure, item analysis and cutting score validation.
The research sample was comprised of four groups of participants that were recruited as couples that were of French Canadian descent. The first general sample consisted of 150 couples that were recruited for a study on marital satisfaction and habitation in the Quebec City region. Two more groups consisted of 60 dissatisfied and 32 satisfied couples, recruited from the same place but for a study using semi open-format clinical questionnaires on the history of their marriage and their current expectations. The final group was recruited for measuring stability and it consisted of 39 couples in the Edmuston region of New Brunswick. All couples included in the study had been together for at least one year.
The measures used in the study were the Locke-Wallace Marital Adjustment Test and a number of other questionnaires about relationships, which are not explained in the article.
The samples in the four groups were all recruited differently. The general sample was recruited by using door-to-door method in two different suburbs of Quebec City Freeston & Plechaty, 1997). They were recruited by having comparable socio-demographic status. The couples were invited to participate in a study on marital satisfaction and living conditions.
The couples were asked to fill out their questionnaires independently without discussing their answers. The experimenter returned a week later to collect the questionnaires
The dissatisfied and satisfied couples were recruited by public announcement based on response to advertisements. The couples were invited to participate in an assessment process that involved a testing period followed by a written and verbal report on their relationship functioning. Test administration was conducted collectively and a research assistant remained present to answer all questions and to assure that no discussion occurred between the partners. A month later the couples met with the research assistants to receive a written and verbal report on the couple’s relationship profile.
The test-retest was recruited among personnel working at the three government institutions in the region. The procedure was identical to the two previous samples with two the excepts.
Reliability is evident through test-retest stability and was measured over a one-month interval. Test-retest reliability for men was .82 and for women .84 when both groups had group sample of 37. For the Hunt scoring reliability was .75 for men and .82 for women. These results of stability are generally considered adequate.
For criterion-Related Validity, the group means of the general, satisfied, and dissatisfied sample were compared using analysis of variance. There were significant differences for both men and women. The dissatisfied men and women scored significantly less than other two samples.
The correlations between partners were all significant and moderate across all four samples. Although the difference was not significant, the correlations were weaker for the dissatisfied couples than for the satisfied couples.
Exploratory factor analysis was conducted on the combined samples. The number of subjects (260 men and 259 women) was adequate for the 15 items. There is a high correlation between factor loadings of solutions for men and women suggesting that the structure is very similar for both.
The LWMAT uses a cutting score of 100. The traditional cutting scored was revalidated using samples of satisfied and dissatisfied participants. The results confirm the pertinence of the traditional cutting-score of less than 100 as indicating maladjustment.
When discussing Item Analysis the relevance of certain items was examined by looking at response distributions. The results showed that some items did differ for men and women.
The Locke-Wallace Marriage Adjustment Test was found to be a well-detailed psychometric examination. The reliability was sufficient. The factor structure was uni-factorial and held constant for men and women for the combined samples. However individual items performed differently for men and women. This suggests that the construct measured may differ for men and women. Ten items discriminated differently for both men and women. Two items may have been scored slightly differently to remove anachronistic concepts of marital adjustment with out any significant effect on the scale of performance (Freeston &Plechaty, 1997). The Satisfied and dissatisfied couples that were recruited for the study offered an acceptable solution to the criterion groups that did not exist in the original Locke-Wallace Marital Adjustment Test (Freeston &Plechaty, 1997). The two groups were recruited in an identical fashion except for their status “Going through Difficult period” and “satisfied with their relationship” (Freeston &Plechaty, 1997).
The test remains appropriate for a clinical context where a rapid assessment is required to assess marriage satisfaction or adjustment. The Locke Wallace Test remains a reliable, valid rapid assessment instrument measuring a broadly based definition of adjustment with only 15 items (Freeston &Plechaty, 1997).
A. Criticism of the test
1. The criticisms of the test relate to scale content, underlying construct, and psychometric issues.
2. Lacks empirical data to support the continued use of the instrument.
B. Role of the test
1. Two important aspects of the test are length and its content
2. There is a need for rapid assessment instruments in clinical and research settings.
1. A measure of martial satisfaction needs to be sensitive to the marital experiences of both genders.
2. Generalizabilty maybe implied by using diverse groups, but cannot be assumed and should be based on separate analyses for men and women.
1. Four groups of couples, all were French Canadian
a. First group equaled150 couples recruited from Quebec City. (General sample)
b. Second group equaled 60 dissatisfied couples
c. Third group equaled 32 satisfied couples
d. Fourth group equaled 39 couples from New Burnswick.
e. All participating couples in the study were required to have been married for at least a year.
1. All received a package including a socio-demographic data questionnaire, the Locke-Wallace Marital Adjustment Test, and questionnaires about relationships.
1. The general sample was recruited using a door-to-door recruitment strategy.
2. Couples were invited to participate in a study on marital satisfaction and living conditions.
a. Confidentiality and anonymity was assured and informed consent were signed before couples received questionnaires.
b. Participants were asked to fill the questionnaire out independently without talking about their answers.
c. The questionnaires were collected five to seven days later by the experimenter.
3. The Dissatisfied/Satisfied sample was recruited by public announcements on the radio, tv, newspapers, and postings in government buildings.
a. Couples were invited to participate in an assessment that involved a testing period followed by a written and verbal report on the functioning of their relationship.
b. First & Third wave announcement invited couples (dissatisfied) that were currently going through difficult times in their relationship.
c. The second announcement invited (satisfied) couples that were satisfied with their relationship.
d. The test administration was done collectively and a research assistant stayed behind to answer any questions.
e. The questionnaire lasted for 90 minutes.
f. A month later participants met with a research assistant to receive a written and verbal report on their relationship profile.
4. Test Retest Sample
a. Sample was recruited from three local government institutions in Canada.
b. Participating couples received a profile after the second test.
A. Reliability is evident through test-retest stability and was measured over a one-month interval.
1. Test retest reliability for men was .82 and for women .84 when both groups had group sample of 37.
2. For the Hunt scoring reliability was .75 for men and .82 for women.
3. These results of stability are generally considered adequate.
B. Criterion-Related Validity
1. The group means of the general, satisfied, and dissatisfied sample were compared using analysis of variance.
a. Significant differences for both men and women
2. Dissatisfied men and women scored significantly less than other two samples.
C. Correlations Between Partners
1. The correlations between partners were all significant and moderate across all four samples.
2. Although the difference was not significant, the correlations were weaker for the dissatisfied couples than for the satisfied couples.
D. Factor Structure
1. Exploratory factor analysis was conducted on the combined samples.
2. The number of subjects (260 men and 259 women) was adequate for the 15 items.
3. There is a high correlation between factor loadings of solutions for men and women suggesting that the structure is very similar for both.
E. Cutting Score Validation
1. The LWMAT uses a cutting score of 100.
2. The traditional cutting scored was revalidated using samples of satisfied and dissatisfied participants.
3. The results confirm the pertinence of the traditional cutting-score of less than 100 as indicating maladjustment.
F. Item Analysis
1. The relevance of certain items was examined by looking at response distributions.
2. The results showed that some items did differ for men and women.
A. The present study concludes that there is still life in test after all.
B. The LWMAT test retest reliability remains reliable, valid, rapid assessment instrument measuring what it says it measures.
C. The test appears to be an empirically adequate measure, but it may not be a conceptually adequate measure.