Introduction
The examination of the functioning of the memory of an individual cannot be limited to only one memory test, and as a result, there are a variety of assessments that target the various features of the memory. Memory test procedures take into consideration various attributes including features of the test instrument such as legitimacy and consistency of results.
Tests should ideally sample a variety of non-related features of the memory to make sure that particular inadequacies in the memory can be identified. In addition to this, aspects of recall and recognition should be tested, as a way to monitor memory processes in relation to memory indicators (Dustman & Beck, 1980).
It has been observed that most of the memory encounters are incidental as opposed to premeditated. In view of this, one of the tests administered should be unexpected, to prevent rehearsal. Some tests are administered through visual aspects, ignoring auditory-verbal presentation modes as well as motor-retrieval response modes.
To make certain that the results obtained are well- founded, the person administering the tests should sample an appropriately broad range of test behaviors. Computerized testing has a number of drawbacks including time consuming and high cost of setting up the system (Dustman & Beck, 1980).
Memory for designs test
The Memory for designs test was designed by Graham and Kendall, in 1946. The test examines visual recall of an individual based on brain damage versus functional disorder versus normality. It is a popular test for the examination of brain damage in children and adults. The scoring system allows the accumulation of normative data (Graham & Kendall, 1947).
Administration of the test
The test is administered by presenting a series of fifteen geometric designs of increasing difficulty on individual cards. The cards are nine by twelve inch pieces of paper, with all designs drawn on the same piece of paper. Each design is presented to the individual alone, in a prearranged sequence. The subject is supposed to view the design for five seconds before it is removed from his view. The test’s duration is about five to ten minutes (Erickson & Scott, 1977).
Development of the test
The memory test was intended to draw the line and distinguish between individuals who are organically impaired and those who are functionally impaired. The inability to reproduce geometric designs from immediate memory is related to organic impairment.
After the test was introduced, forty designs were set and administered to a few individuals who had brain damage. The final set of tests included those designs that were found to be easiest to score as well as discriminate best between groups. The test has since been used with all age groups for both clinical and research purposes (Erickson & Scott, 1977).
Sample
According to Graham and Kendell (1960), there were a variety of samples such as the original validation group and a cross-validation group. In the original group, there were 70 brain-disorder patients with mixed diagnoses and 70 controls. They were all matched for age, education and occupational status.
The original group included an equal number of males and females, as well as African Americans. The age bracket of the subjects was 9 to 69 years, with an average age of 42 years. The average educational level was eighth grade. The latter group comprised 33 brain-disordered patients and 168 control patients. There were three times as many men as women, with an average age of 28 years for the controls and 37 years for the brain-disordered subjects. Both groups had an average education of ninth grade (Peterson & Mangen, 1981).
Scoring, scale norms and distribution
According to the criteria provided by Graham and Kendall (1960), each reproduction has a score of 0-3, whereby the highest score signifies the worst performance. There was no penalty due to incomplete or forgotten designs. This is because the variables did not distinguish brain damaged subjects from control subjects.
Reliability tests
According to Graham and Kendell (1960), the split half reliability was .92 for 140 brain-damaged patients. Test-retest correlations conducted in the same session or within a 24 hour period for other groups ranged between .81 and .90. The clinical usefulness of the test is based on the adequacy of its standardization for various age groups as well as a variety of non-brained-damaged, psychiatric individuals.
The reliability of the scoring is suitable based on data provided in the test manual. In addition to this, new studies on the reliability have found it satisfactory. Qualitative scores are possible based on a measure of the tendency to rotate the designs, whereby the error can be reliably calculated (Graham & Kendall, 1947).
The age standardization for normal adults is adequate though the normative data that is available for different functional psychiatric groups is not. It has been observed that some functional psychiatric disorders produce lower scores on the MFD test, though there are no age norms for representative populations of functional psychiatric disorders.
This poses a challenge in the omission of the presence of a specific functional psychiatric disorder as a substitute explanation of a poor score based on the presently available norms, unless the alternative was eliminated on a different basis (Peterson & Mangen, 1981).
Validity tests
An analysis of a large number of tests conducted on a sample of a hundred people to diagnose brain damage showed a +.597 for the memory for designs test on the first factor. This was interpreted as perceptual organization or the ability to integrate the relevant aspects of the perceptual field. Based on the perceptual organization factor, a memory aspect was observed from the tests, proving that perceptual organization was able to facilitate performance when memory was involved (Peterson & Mangen, 1981).
The MFD test has been proven to always differentiate between groups of patients known without any doubt to have brain damage, as well as to match groups of normal people and psychiatric patients thought not to be brain damaged.
According to the definition of brain damage, that it is any amount of cell death in the cerebral cortex irrespective of the cause, every person aged above 30 years would be brain damaged. Some forms of birth trauma, childhood injuries and anoxia cause cell death in many normal children, though the damage is insignificant since it does not result in impairment (Richie & Butler, 1964).
Predictive validity
According to data presented in the Graham and Kendall ( 1960) test manual, a significant mean score difference between a group of brain-damaged and normal individuals, at the 0.01 level. This indicated an overlap in the scores of the non-brain-damaged and brain damaged groups that caused difficulties in coming up with a decision on the performance of borderline people.
An example was the cut-off point whereby 4% of the controls were diagnosed as having brain damage. Out of these, half of the brain-damaged subjects in one group and 48% in another group were designated with brain damage. The other cut-off point made for varied combinations of correct identifications, errors and false positives (Richie & Butler, 1964).
Convergent validity
According to Warren and Mangen, the memory for designs test correlated .85 and .81 with scores on the Bender Gestalt test.
Usability on older populations
The test is timed. In addition to this, manual coordination is required. The involvement of these factors implies that the aged may be handicapped.
Sensitivity to age difference
According to Peterson and Mangen, there was a significant correlation between performance and age, of 0.27. This implied that an increase in age of the subject corresponded to poorer performance of the subject. There was also a correlation of -.31 of the MFD test scores with intelligence. Assessment by the vocabulary score on the Wechsler-Bellevue showed a significant multiple correlation of .44 with age and intelligence (Peterson & Mangen, 1981).
According to a study conducted on psychiatric patients aged above 60 years, 36 brain damaged and 36 controls, there was a considerable difference in the mean scores, which was not as high as that between, brain-damaged and standardization groups of younger ages.
The mean scores were 8.44 and 13.89, while the SD was 6.33 and 5.81 for the controls and brain-damaged group respectively. There was a big difference between this age group and all younger control groups, though the difference with the scores of adults between the ages of 20 and 60 did not vary much.
The confounding factor was that the controls over 60 years had a decline in vocabulary scores that was the same as that of the brain-damaged groups, which yielded negative results for the younger ages. This signified that the use of psychiatric controls for the older age-group would be unsuitable (Dustman & Beck, 1980).
According to another study conducted for subjects between the age of 20 and 80, with 50 men and 40 men in each constituent decade, the performance of older age groups on the MFD was observed to be stable until the age of fifty, after which there was a fast decline and increased variability in subjects’ performances (Peterson & Mangen, 1981).
Conclusion
The test’s reliability and validity are good, since it discriminates in a rational manner between groups with brain damage and normal groups. The test has limitations when conducted on aged subjects since it requires manual coordination. In addition to this, it is timed and the performance correlates with age and intelligence.
Unusual scores signify damage in a specific cortical area, which in turn signifies a formerly particular dysfunction. It could be translated to mean a shortened life expectancy, though the MFD is hardly translated to imply such occurrences. Studies have shown an insignificant correlation between MFD score and a rating of severity of brain damage according to certain assessment criteria such as EFG (Peterson & Mangen, 1981).
Other studies have also shown a wide range of MFD scores when the test is administered to a psychiatric group whose brain damage status is uncertain, resulting in no conclusions from the features shown by the group. One practical consequence of an abnormal MFD score which has been demonstrated is that the patient is likely also to have abnormal scores on several other tests of brain damage such as the Bender Gestalt Test, and the Benton Visual Retention Test.
An abnormal MFD score might enable the clinical psychologist to identify a previously unknown pattern of associated psychological abnormalities which might have implications for the adjustment of the patient. Another useful implication of an abnormal MFD score among children is that one might anticipate difficulties in learning to read, as suggested by several studies (Erickson & Scott, 1977).
References
Dustman, R. E., & Beck, E. C. (1980). Memory-For-Designs Test: comparison of performance of young and old adults. Journal of Clinical Psychology , 36(3), 770-774.
Erickson, R. C., & Scott, M. L. (1977). Clinical Memory Testing. Psychological Bulletin , 1130-1149.
Graham, F., & Kendall, B. (1947). Memory-For-Design’s Test. Journal of Consulting Psychology , 11(6).
Peterson, W. A., & Mangen, D. J. (1981). Research Instruments in Social Gerontology: Clinical and Social Psychology (Clinical & Social Psychology). Minneapolis: University of Minnesota Press.
Richie, J., & Butler, A. (1964). Performance of retardates on the memory-for-designs test. Journal of Clinical Psychology , 20(1), 108-110.