Figure Drawing Techniques with Older Adults: A 30 Year Perspective
Paul E. Panek, Bert Hayslip, Jr., Sharon Rae Jenkins and Amanda Kay Moske
Panek, Wagner and Kennedy-Zwergel (1983) and Hayslip and Lowman (1986) have reviewed the use of expressive personality tests with older adults. In this context, the present review presents and discusses the research on human figure drawing techniques (HFD) with older adults during the past 30 years since those reviews. Findings are presented for age differences, and the differentiation of diagnostic groups.These are discussed in the context of the assessment referral questions most common with older adults and the challenges of geriatric assessment for which HFDs are suitable tools. Conclusions and recommendations regarding the use of HFD with older adults are presented, as are the implications for future research using HFD with older persons.
Psychologists have recognized the importance of becoming skilled in working with older persons, especially those with some form of dementia (Abeles et al, 1998; APA Monitor, 2008; Panek, 2006). However, many clinicians have little professional training in working with such persons (see Knight, 2004), for whom certain referral questions, differential diagnoses, and appropriate assessment tools may differ from younger populations. Psychological assessment instruments are available to assess diagnosis, levels of impairment, personality, and social functioning, as well as to distinguish premorbid personality and normative age differences from psychopathology and responses to situational stress. Such assessment information is needed to inform decisions affecting older adults and their families, for example the wisdom of independent living versus an assisted situation versus admission to a long-term care facility, level of care needed, the older person's capacity to cope with the emotional byproducts of physical or cognitive deficits, or the implementation of treatment plans and other interventions, including the likelihood of treatment adherence. However, these
instruments do not necessarily address these most common referral questions for this population, and they are not always suitable for older adults for reasons discussed below.
We present here a review and evaluation of the empirical and clinical literature regarding the status of human figure drawing (HFD) techniques with older adults over the past 30 years since the reviews by Panek, Wagner and Kennedy-Zwergel (1983) and Hayslip and Lowman (1986). These are classified as constructive techniques (see Meyer & Kurtz, 2006 and Radika & Hayslip, 2002). We hope to provide an objective basis for clinicians' decision-making regarding whether to utilize HFD techniques either formally or informally in their evaluation of older clients. We begin by presenting the major clinical referral questions for which HFDs may be uniquely useful with older persons, then we discuss broadly the suitability of HFD for addressing those questions in this population. The major portion of the paper reviews the past three decades of literature on HDFs with older adults, and is followed by our synthesis, conclusions, and recommendations.
……………………………………………………………………………………………………………… Bert Hayslip, Jr., 821 Sail Lane #101 Murrells Inlet, SC 29576, Email: Berthayslip@my.unt.edu. Regents Professor Emeritus, Dept. of Psychology, University of North Texas, Denton, TX, USA (Correspondind author), Sharon Rae Jenkins, Professor of
Psycho, Univ. of North Texas, Denton, TX, USA and Amanda Kay Moske, Dept. of Psychology, University of North Texas, Denton, TX, USA.
Key Words: Expressive Personality
Note: This paper is dedicated to the first author, Prof. Paul E. Panek, Dept. of Psychology, The Ohio State University-Newark, who died on December 28, 2007.
Common Assessment Referral Questions with Older Adults:
As the proportion of older persons in our population increases, the likelihood of such persons experiencing certain forms of mental illness also increases (Hayslip, Patrick, and Panek, 2011; Koenig, George, & Schneider, 1994). Depressive symptoms, anxiety, and dementia are the most likely forms of psychopathology that clinicians are likely to confront, wherein the incidence of subclinical depressive symptoms and dementia are not uncommon in later life (see Segal, Qualls, & Smyer, 2011; Zarit & Zarit, 2008). Certain subtypes of non-melancholic depression may be secondary to some older adults' living situations (e.g., in nursing homes, socially isolated, physically inactive) and/or their anticipation of resource depletion, decline, and d e a t h ; c o n s e q u e n t l y, t h e y m a y g o unrecognized and underdiagnosed (see Blazer, 2002; Zarit & Zarit, 2008). Whether or not these symptoms reach the threshhold required for formal diagnosis, they may exacerbate social isolation and inactivity, contributing eventually to physical decline. Confronting and coping with such mental health challenges is essential to older persons' and their families' quality of life.
Routine stressors associated with aging include age-related sensory losses, losses of close others as family and friends die, losses of functionality, and eventually loss of self. Anticipating and experiencing these demands may elicit depressive reactions and anxiety about the future (see Ruth & Coleman, 1996; Segal,et al, 2011; Zarit & Zarit, 2008). Older adults may be coping with the physical and cognitive changes associated
with aging as well as changes in family relationships growing out of such changes (see Blieszner & Bedord, 2012). Some older persons experience significant difficulties in dealing with the normal physical changes accompanying the aging process, i.e. sagging and wrinkling of
skin, loss of height, redistribution of fat from the extremities to the torso, which impact persons' assessment of their physical attractiveness as well as their physical competence (Whitbourne & Skultety, 2002). These changes may require identity assimilation or identity accommodation as an aspect of coping to retain a coherent sense of the self despite physical change. Many older persons (and their families) must face the ultimate loss of the self that is associated with Alzheimer's disease, stroke, or other forms of dementia (Feil, 2002; Taylor, 2007).
Declines in one's physical or cognitive functioning may have a unique psychosocial impact on older persons' personality and behavior secondary to the actual functional losses. For example, illnesses such as osteoporosis, arthritis, or cancer (see Fiedler, 2012; White, 2002) as well as the effects of cancer treatments themselves (Pruzinsky, 2002) may impact physical competence, undermining older persons' self-efficacy. This may impede older adults' efforts to engage actively in their prevention, maintenance, and/or treatment regimens to prevent or remediate the effects of normal and pathological aging changes (e.g. the use of sunscreen, diet, exercise; Whitbourne & Skultety).
Psychometrically measured changes in one's functioning may be a byproduct of the older person's reaction to the loss of his/her skills (e.g., depression symptoms, feelings of inadequacy, sense of losing oneself, impaired relationships with others, withdrawal from others, anger, changes in sleep or patterns of eating), which tends to lower a person's scores on timed tests. Constructive test performance is not a direct alternative to psychometric measures of cognitive functioning, but with expert interpretation, it may assist in diagnosis by providing evidence for or against specific DSM criteria, or appraising the client's status on rule-out criteria (see Segal, Qualls, &Smyer, 2011; Zarit & Zarit, 2008).
One's reactions to the immediate life context (e.g., chronic health difficulties and resulting treatment burden, social isolation, an understimulating nursing home environment, exposure to ageism in everyday interactions, difficulties in everyday decision-making) may present additional stresses and/or deprivations that tempt misattribution of situational stress effects to psychopathology. Reichlin (1984) notes that ignorance of the impact of situation-related moderating factors on the test performance of older persons may lead to their being overpathologized by the clinician. Complicating matters is the increased range of individual differences in functionality with aging (see Nelson & Dannefer, 1992), associated psychosocial and health problems, and psychological responses to these that may interfere with treatment adherence or result in older adults requiring a variety of support services, adding to the range of referral questions and recommendations. Like verbal expressive techniques (see Panek, Jenkins, Hayslip, & Moske, 2013), constructive techniques to assess personality such as the HFD techniques can play a unique role in informing these recommendations in the context of multimodal assessment.
Why Use HFD Techniques for Assessing Older Adults?
Psychological assessment with older adults faces unique challenges related to American cultural expectations regarding aging and its consequences, and the familial, social, and institutional structures that define and shape the aging process. The co-morbidity of either cognitive or physical change and personality difficulties, along with situational stressors, complicates the usual purposes of assessment and adds a cascade of new ones that are less typical of younger individuals. In addition to the stressors described above, older adults may fear losing the ability to live independently if impairments are discovered. This fear may
make some resistant to clinicians' efforts at diagnosis by explicit self-report, structured interview, or psychometric methods. On the other hand, some isolated, lonely older adults who are anxious about their health status may overengage with clinical interviews and other less structured self-reports, having learned to use elaborate symptom disclosures as a form of social discourse. Either orientation is likely to give misleading results. Memory impairments may make self-report findings questionable. Physical impairments might make some assessment methods too fatiguing to yield accurate results, leading to overpathologizing the client. Subclinical depression symptoms may slow performance on timed tests, yielding underestimates of functional capacity. Word-finding problems may constrain scores on verbal expressive tasks, leading to underes-timates of language comprehension. Reliance on DSM diagnosis (except in extreme cases) tells the clinician little about the client's ability to sustain independent living and does not address explicitly the kinds of referral questions for which older adults are often referred for psychological assessment.
Human Figure Drawing (HFD) techniques are sensitive to subtle personality changes that often precede the diagnosable symptoms of the Alzheimer's form of dementia, i.e. Mild Cognitive Impairment (Peterson et al., 2001). HFD techniques may provide us with valuable information about how the older person is coping with either alterations to their bodies via aging or disease, or normal or pathological cognitive decline, emphasizing their strengths in doing so (see Kahana, 1978). Moreover, for some older persons, they can be less threatening than clinical interviews, more acceptable and engaging and less fatiguing than are lengthy personality inventories, and may allow for the detection of subtle response patterns that more structured and self-report measures obscure.
We note eight key points briefly. First, HFD
techniques are routinely used by practitioners as clinical tools, although their perceived utility covaries with the skills and experience of the clinician utilizing it (Anastasi & Urbina, 1997; Handler & Thomas, 2014; Qualls & Knight, 2006; Segal, Qualls, &Smyer, 2011). Second, techniques such as HFDs do not require verbal skills or insight, and are easily adapted to assess older persons who are embarrassed or guarded about what they might reveal about themselves in the process of being evaluated (Anastasi & Urbina; Kane & Kane, 2000; Segal et al.). Third, HFDs may help in the establishment of rapport and easily engage older persons (Radika & Hayslip). Fourth, HFDs may provide data that may be more intuitively obvious to older patients and their caregivers than psychometric scores when used with collaborative (Fischer, 1994) or therapeutic (Finn, 2007) assessment approaches. Fifth, HFDs are faster and easier to administer and score than comprehensive structured interviews and personality inventories, and thus may be useful for screening (REF). Sixth, it could be argued that HFD techniques are best suited to be used as an adjunct to either structured or unstructured interviews and other assessment tools, key to the broadband assessment of older persons (see Anastasi & Urbina, 1997). Seventh, process information gained from qualitative clinical tools such as HFD techniques at times can be more important that quantitative data in the wholistic assessment of the older person (Abeles et al., 1998). Finally, they may indeed be valuable in generating hypotheses about the client, based upon the characteristics of the drawings themselves and an inquiry following the drawing task, which must be interpreted in the context of other information about the individual (Anastasi & Urbina; Zarit, Eiler, &Hassingher, 1985). Edelstein and Segal (2011) echo this preference for multimethod assessment in discussing the shortcomings of relying solely on self-report data and suggesting that the shortcomings of one method can be avoided to an extent by relying upon multiple methods of
assessment and triangulating the evidence.
A last key consideration is that the utility of the criteria by which constructive techniques are evaluated is a function of whether or not they are considered tests, and thus subject to the psychometric criteria used to judge structured self-report scales and performance measures. Alternatively, they may be used and interpreted as process-based measures, as is done by most clinicians (see Jenkins, 2008a). In this respect, Hooker and McAdams (2003) and Bornstein (2009) stress the unique value of a process approach for assessing not only personality traits, but also goals, motivations, self-regulatory processes, expectations, and the construction of life stories. Anastasi and Urbina (1997) observed, regarding “projective” techniques, that “attempts to evaluate them in terms of the usual psychometric procedures may be inappropriate” (p. 441). In a similar vein, Neugarten (1977), in discussing late life personality, questioned whether personality can be meaningfully understood in solely scientific terms.
On the other hand, Zarit, Eiler, and Hassinger (1985, pp. 736-737), in discussing the “projective” assessment of personality in later life, note that “at the heart of the matter is the extent to which untested and inferred propositions are appropriate in assessment
…….Inference, intuition, and creativity all have an important place in assessment as well as other clinical activities, and they are useful for generating hypotheses to account to account for the apparent diversity or order in human phenomena, and it is often necessary to go beyond the data, if only because adequate explanations are lacking. At the same time, inferences that are supported only by other inferences can be misleading”. Thus, the use of scoring manuals that are sufficiently detailed to allow for appraisal of interscorer reliability is the most important indicator of objectivity.
We argue that at the minimum, HFDs should be
viewed as an adjunct to other sources of information the clinician has at his/her disposal regarding the older client. Interestingly, in a study of 158 British clinical psychologists, while most did not utilize figure drawings as a formal component in their assessment, half reported using drawings as an informal aid in assessing persons (Bekhit, Thomas, & Jolley, 2010).
Purpose and Structure of the Present Review :
The present review, like its predecessors, discusses the research with HFD techniques with older adults during the past 30 years, organizing the review by 1) the extent of age differences in performance and 2) diagnostic group comparisons. While too few studies address similar questions with the same outcome measures to allow for a meta-analysis, in discussing this work, we provide relevant sample characteristic data and information pertinent to effect sizes (e.g., Cohen's d, effect size r between the independent and dependent variable, r2) where provided and/or where sufficient information was provided by a study's author(s) enabling us to compute effect sizes. Where work has been replicated or cross validated, we note this as well. Where group comparisons were conducted, we note whether they were systematically matched to control extraneous variables, and if they were not explicitly matched, whether their demographics were comparable, or whether the comparability of groups was unclear.
Overview of the HFD Research Literature over 30 Years:
Since the earlier reviews by the first and second authors, drawing techniques have become the most researched expressive personality tests with older adults (see Radika & Hayslip, 2002). The main foci of these studies have been to establish normative age differences, to differentiate Alzhiemer's disease and other dementia patients from normal controls, and/or to identify levels of personality that may parallel
different levels of cognitive impairment. HFD techniques are also a prime source of data in the study of body image, but this research has largely neglected older adults, as noted by Peat, Peyerl, and Muehlenkamp (2008). Since the previous reviews, no new work regarding either individual differences or drawing techniques' normative characteristics with older persons has been published, so these sections from the earlier reviews are not presented here.
Age Differences:
In a study by Ericsson, Hilleräs, Holmén and Winblad (1996), using a freehand HFD technique (FFD, where persons were simply instructed to draw a human figure) in a sample of 668 persons aged 75 and older, the authors found that age had a significant influence on HFDs after age 90, wherein most HFD variables showed regressive changes; this was true as well for persons suffering from dementia (see below). A reduction of essential bodily details with increasing age (and level of impairment), based upon a cross sectional (10 groups of 10 five-year cohorts) comparison of persons aged 35-80, has been observed by Ericsson, Winblad and Nilsson (2001).These authors argue the human figure drawings are as good a predictor of cognitive declines in older persons as is episodic memory performance, and consequently can be used with confidence to screen older persons for cognitive decline; among the oldest aged persons, this may be especially valuable (Ericsson et al.).
Ericsson, Hilleräs, Sundell and Winblad (1997) examined HFDs in a) 716 children/adolescents who were enrolled in a longitudinal study; b) school pupils (n = 143) between the ages of 10 and 15, c) nursing students (n = 71) aged 20 to 35, d) 936 healthy adults aged 35-80, e) 371 elderly living in the community, and f) 94 cognitively impaired elderly, as determined by mental status exam scores and clinical dementia rating scale performance. Figure drawings were evaluated for essential body details (e.g. facial features, hands, legs),
hierarchical level (objectively scaled childlike qualities or not, e.g. drawing stick figures or just a head and feet), and structural characteristics (centeredness regarding placement on the paper). One-way ANOVAs suggested no age differences between the ages of 10 and 80 years for the presence of essential details, while hierarchical levels (drawings that were more technically correct, i.e. body parts are not just present, but are drawn in correct proportions relative to one another) of the HFD decreased with increasing age. That is, in the youngest (4 to 5 years of age), drawings changed from being defined by essential details to correct proportions. At 70 years of age, drawings continued to reflect correct proportions, while from 75 to 90 years of age, figures were drawn in two dimensions. Thus, few age differences were observed, except between the very young and the very old. Regarding the two older adult groups, impaired aged persons (age > 75) as a whole group showed, like the 4-year old group, a large range in drawing essential details. Of the structural characteristics, centeredness varied in all groups except nursing students and impaired older persons. This study's principle strength is the diversity of the samples varying in age. Yet, it is cross sectional in nature, and consequently cannot speak to intraindividual change over time. It is also to be noted that many of the age effects, while reliable, are nevertheless small in magnitude.
In a study by Hayslip, Cooper, Doughtery, and Cook (1997), human figure drawing scores were reliably derived according to the system developed by Doughtery and Cook (1991), and broken down into general characteristics, e.g. figure size, line quality, and specific characteristics, e.g. eyes, ears, arms, feet. Such scores favored young adults (ES rs> . 14). Thus, figure drawing scores converged with Holtzman Inkblot Technique (Holtzman, Thorpe, Swartz, & Herron, 1961) scores in suggesting that older persons exhibited a more negative body concept at both a global and a specific level than middle-aged and younger
individuals, and indicate that body image issues may indeed exist for middle-aged and older adults, not just younger adults (Hayslip et al.).
Silver (1987) investigated the effects of gender and age on the expression of emotions using the Stimulus Drawing Task (Silver, 1986), comparing third graders, high school seniors, adults (between the ages of 20-30 years), and older adults (total N = 326). Silver found gender differences in the emotional content of the drawings across the four age groups (F = 4.15, p< .05, ES r= .22). To the extent that the principal subject of a drawing reflects the self-image of the person who draws it and the environment reflects the way that person perceives the world, these findings suggested that men and boys tend to see themselves as fighting in a dangerous world, while women and girls tend to see themselves as part of the world rather than opposing it (Silver). Given the limited demographic information regarding participants' age and health in this study, it is difficult to assess the significance of the lack of age and presence of gender differences found by Silver here.
Silver (1993a) explicitly examined age and gender differences in attitudes toward solitary subjects and relationships expressed in response to the Stimulus Drawing Task (Silver, 1991). Participants were 531 respondents in five age groups: children (7-10 years), younger adolescents (13-16 years), older adolescents (17-19 years), younger adults, (20-50 years), older adults (65+ years). Results indicated: a) respondents tended to choose and draw subjects of the same gender as themselves, ÷²
= 145.84, p< .001, ö = .66; b) boys and men expressed positive attitudes toward solitary subjects, negative attitudes toward relationships, ÷² = 46.97, p< .001, ö = .47; c) girls and women expressed positive attitudes toward solitary subjects, both positive and negative attitudes toward relationships, ÷² = 25.32, p< .001, ö = .33; and d) boys and men showed higher frequency than girls and women in drawing about assaultive relationships, ÷² =
9.38, p< .01, ö = .21. However, age and gender interacted resulting in a significant age variability in assaultiveness for females (÷² = 11.89, p < .05, ö = .24) but not for males. The proportion of older women who drew pictures about assaultive fantasies (27%) somewhat exceeded both the proportion of older men who did so (21%), and the proportion of all other female age groups. An inverse age by gender interaction was also found for caring relationships (÷² = 12.52, p< .05, ö = .25), where males showed significant age variability (÷² = 13.10, p < .05, ö = .25), while females had a significant commonality of caring relationships across all age groups. The proportion of younger men who drew pictures about caring relationships (46%) exceeded both the proportion of younger women who did so (10%) and the proportion of all other male age groups (Silver). Silver suggested that drawings can identify age and gender differences in attitudes, and if expanded, they may eventually provide norms upon which to base more accurate expectations as well as provide information about emotional needs.
Silver (1993b) expanded on Silver (1987) by more closely investigating attitudes toward self and others. One group of participants was 59 adults over the age of 65 who lived independently in their communities in New York City and who varied in socioeconomic status (moderate to low SES versus high SES) who responded to the Stimulus Drawing Test (Silver, 1991). The second group consisted of 22 older adults from moderate to high socioeconomic neighborhoods in a Florida city, responding to the Drawing from Imagination subtest of The Silver Drawing Test of Cognitive Skills and Adjustment (SDT). Despite the fact that samples were too small for statistical analysis, the authors nevertheless claimed that among older adults there was expressed pessimism in the subjects they chose to draw and the relationships they portrayed. Silver infers that these findings are consistent with the idea that depression may be widespread among older
men and women in our society; this is indeed true regarding depressive symptoms but not major depressive disorders (see Segal, Qualls, & Smyer, 2011; Zarit & Zarit, 2008).
Morin and Bensalah (1998) found in persons aged 30-85 that self-portraits differed with regard to the representation of the hands and face, where among persons younger than 65, no one failed to represent facial features, while 17% of persons aged 65 and older were unable to produce self portraits with facial details; this was interpreted by the authors as reflecting the inability to face the losses accompanied aging, given that the drawn real mouth and eyes are both symbolic body orifices and imaginary in nature. In a sample of 225 persons aged 3-80 who completed a drawing completion task where instructions were to draw simple lines suggesting 2 figures facing one another, Spaniol (1992) found the lines of older adults to be more likely to draw broken lines, with less differentiation and fewer boundaries; this might suggest the perception of less clearly defined relationships and more dependency in later late. Pruis and Janowsky(2010) presented younger and older women (ns = 19 in each case) with morphed images of their own bodies (ranging from 35% thinner to 35% fatter than each woman's actual size). Such ratings did not differentiate younger and older women, suggesting equal degrees of body satisfaction with age.
Comparing Diagnostic Groups :
Relative to other expressive techniques, much HFD work has appeared regarding the comparison of diagnostic groups, especially cognitive impairment and dementia, suggesting that older persons can be successfully differentiated in this respect utilizing HFDs. To explore figure drawings' ability to differentiate older persons on the basis of cognitive impairment, Ericsson, Forssell, Amberla, Holmen, Viitanen and Winblad (1991) administered several drawing/graphic tasks (free-hand figure drawing, FFD), and the three-
dimensional Cube Test to 250 older (aged 75 and older) persons with cognitive deficiency (according to mental status exam scores) who were matched on age and gender with unimpaired individuals. Scoring of the FFD reflected the presence of eight critical details (i.e., main details such as two-dimensionality of arms and legs, clothing, body proportion, shoulders, ears, movement and eyes turned in same direction). The number of main details disappeared in fixed order with decreasing cognitive function. The authors observed that the order of loss of these details was the inverse of that observed during the development of the child. Overall, declines with greater cognitive impairment were observed for all variables (p<
.01 for ears and life-like movements, p< .001 for clothing, arms and legs in two dimensions, body proportions and shoulders; no F values were reported) with the exception of directed glance. The authors suggest a subset of the different tasks of copying geometrical figures, writing sentences, FFD, and writing a signature can be used to evaluate an individual's state of cognitive functioning.
Knapp (1994), using modifications (M) of the House-Tree-Person (H-T-P-M) and the Kinetic Family Drawing (K-F-D-M), compared early stage Alzheimer's disease patients (n = 25, aged 51-89) and an age-gender matched healthy control group (n =25, aged 45-83) . These drawing tests differentiated persons by group (t = 11.45, p < .0001, d = 1.64), though the most substantial difference in this respect was in H-T-P-M scores (t = 12.57, p< .0001, d = 1.80). Knapp suggested these results demonstrated the effectiveness of expressive drawing tests in differentiating pathological and normal aging processes. It is of note that groups were matched on age and gender here, but the samples of normal and impaired persons were small, and no controls for drawing competence were implemented.
Ericsson, Hilleräs, Sundell and Winblad (1996) examined freehand figure drawings (FFDs) in
465 elderly people aged 75 and older, 94 of whom were cognitively impaired. They were matched on the basis of age with a control group of 371 cognitively healthy elderly living in the community. Figure drawings were evaluated for essential body details, hierarchical level (e.g. non-response, scribbles, circle, headfooter, stick figure), and structural characteristics (e.g. ratio of head size to figure height, centeredness of the figure on the page). Findings indicated that the number of body details and figure height decreased with decreasing cognitive function, whereas centeredness increased with decreasing cognitive function. However, the FFD did not discriminate between individuals with Alzheimer's disease and persons with vascular dementia. While no effect sizes could be derived from this study, Ericsson et al. suggested that the FFD can help differentiate elderly persons with and without dementia as well as differentiating persons with dementias of varying severity. It is to be noted that some, but not all of the effects found in this study were small, though reliable, a likely artifact of the increased power associated with large samples. This study's principle strength is its large and representative sample and the differentiation of essential body details, attention to developmentally relevant drawing characteristics, and structural characteristics.
Ericsson et al. (1994) compared the free-hand human figure drawings (FFD) of mild dementia patients (n = 11), those with moderate dementia (n = 22), and those with severe dementia (n = 27); all groups averaged between 78 and 82 years of age. Of the dementia patients, 36 were diagnosed with Alzheimer's disease (AD) and
24 diagnosed with Vascular Dementia (VD). The FFD was scored for 53 body details according to Goodenough's Draw-A-Man test, but modified for older persons (Ericsson et al., 1991). A short scale of 15 details was developed by selecting body details with high item-total correlations (SHFD). The SHFD had high interscorer (r = .96) and test-retest reliability (r =
.85) and high concurrent validity in that the SHFD was highly correlated with the MMSE scores (r = .78) and the Clinical Rating Dementia Scale (CDR) (r = .82). A one-way ANOVA indicated that the SHFD discriminated demented and non-demented patients and different levels of dementia severity (F = 82.6, p< .0001, ES r = .80); no differences were observed between AD and VD patients. Ericsson et al. suggested that the HFD is a valid instrument for assessing cognitive dysfunction in old age. While their two studies suggest that the HFD can differentiate older persons by level of impairment, their samples are very small. A similar degree of impairment (spontaneous drawing and copying a figure) among 15 dementia patients versus 15 age and gender matched healthy controls was found by Moore and Wyke (1984), though no information of the strength of these differences was provided by the authors.
Ericsson, Hilleräs, Holmén and Winblad (1996) studied the effectiveness of freehand human-figure drawings (FFD) (persons were simply instructed to draw a human figure) in a sample of 668 persons aged 75 and older to differentiate elderly people with dementia (n = 444) and healthy elderly controls (n = 224). Figure drawings were analyzed via the content of body details and structural characteristics of the drawings; findings indicated that the body details and the height decreased with decreasing cognitive function, whereas the centeredness increased with decreasing cognitive function. However, the FFD did not discriminate between individuals with Alzheimer's disease and persons with vascular dementia. While no effect sizes could be derived from this study, Ericsson et al. suggested that the FFD can help differentiate elderly persons with and without dementia as well as differentiating persons with dementias of varying severity.
Wang, Ericsson, Winblad and Fratiglioni (1998) explored whether the screening capacity of the
MMSE (both sensitivity and specificity; see Kane & Kane, 2000) could be increased by adding the HFD test, comparing data from 95 older persons who were affected by dementia and 366 who were non-demented. Controlling for HFD scores, MMSE scores were found to be significantly lower in the groups with different severities of dementia (p< .001) versus the non-demented group. Moreover, scores from a short version (29 HFD body details, HFD29) and seven HFD essential body details (HFDess) were significantly decreased in the demented as compared to the non-demented group (p<
.01, d = 1.29). Wang et al. found that sensitivity increased by 4.2% when HFD and HFDess were added to the MMSE, compared to MMSE scores alone, concluding that the HFD29 can complement information gathered from the MMSE, especially in the relatively early stages of dementia (corresponding to Mild Cognitive Impairment, MCI; Peterson et al., 2001).
To explore visuoconstructive deficits in impaired versus unimpaired older adults, Freeman et al. (2000) asked 63 older adults suffering from either Alzheimer's disease (AD), vascular dementia, or Parkinson's disease (versus normal controls, n = 20) to copy a modified version of the Rey-Osterrieth Complex Figure (a figure similar to two superimposed pentagons derived from the Bender-Gestalt test). As might be expected, normals outperformed all dementia patients; and the drawings of AD patients were of higher quality (less fragmented, fewer omissions and perseverations) than those of the other demented groups; this finding was quite robust (d = 1.9). Similarly, Fan, He, Xiao and Jia (2002) found that the HFD performance of persons with mild Alzheimer's disease exceeded (p <
.05) that of persons with moderate AD, and that all persons with AD performed more poorly than normal; body details differentiated persons with different degrees of cognitive impairment as measured by the Clinical Dementia Rating Scale. However, these authors found no HFD differences in comparing AD and vascular
dementia patients, using the Draw-A-Man test. Helmes (2012) found a similar difference in comparing older persons who were unimpaired, suffering from dementia, or presenting with some cognitive impairment, but without dementia in terms of their ability to copy the intersecting pentagons from the Bender-Gestalt (rated along a 10-point scale). As did Freeman et al., Helmes found those who were unimpaired to outperform those without dementia, who in turn, outperformed those with
qualitative assessment (DAP 50 - based on five components from examination of the drawings; each component has a 10-point value). Participants were members of one of three groups: Dementia (n = 51) at three levels of severity; Depression (n = 70), and Controls (n = 39). Participants were similarly aged, but were not matched by gender. However, DAP scores were not related to gender, education, or age (p
= .14). Mann-Whitney U rank tests indicated that DAP scores of individuals with dementia
dementia (R2
= .26, d = .90). In contrast,
differed from those of both controls (U = 460, p <
(Jungwirth et al., 2009) found that when combined with purely cognitive measures (verbal fluency, episodic memory), figure drawing performance did not adequately predict the incidence of late stage Alzheimer's disease (AD) in a longitudinal study of older adults (average age = 78, n = 478), suggesting that, non surprisingly, such measures better reflect deterioration per se characteristic of AD, wherein many cognitive measures are sensitive to such declines (see e.g., Potter & Attix, 2006).
A major problem in the clinical diagnosis of older adults is to differentiate depression and dementia, since many of the symptoms of depression mimic dementia (Wells, 1979). Thus, Clément et al. (1996) assessed the quantitative and qualitative differences characterizing human figure drawings and comparing depressed, demented, and healthy elderly persons in two studies. In the first study, the DAP was given to 39 dementia patients and a matched (gender, age, education) control group of 41 older adults. Mann-Whitney U rank tests indicated group differences in DAP scores (U = 237, p< .001), as well as for the MMSE (U = 57, p< .001). Subsequent analyses indicated good inter-rater reliability for the 20 scored items (Kappa > .65, p< .05). These 20 items made up a second (new) quantitative scale with an internal consistency of .95 (Cronbach's alpha).
In Study Two, Clément et al. (1996) administered the new 20-item DAP and added a
.0001, ES r= .50) and those with depression (U
= 1,057, p< .0001, ES r= .36). However, there were no differences in DAP scores between controls and those with depression (p = .19). Subsequent analysis indicated that four items demonstrated good sensitivity in discriminating normals from those with dementia: presence of: eyes, nose, arms, and the connection of the arms to the trunk (p< .001). The presence of arms, the connection of the arms to the trunk, and the presence of nose, hair, ears and the trunk length discriminated between dementia and depression patients. Only one item, eyes present, discriminated between the controls and depression patients (p = .007). Clément et al. (1996) suggest that the drawings of older a d u l t s a r e o f t e n c h a r a c t e r i z e d b y incompleteness, a lack of integration, flatness, a lack of proportion, bizarreness, evidence of inadequate motor coordination, and also make it difficult to identify the gender of the figure. These authors recommend caution in the use of the DAP with older adults due to a problem with inter-scorer reliability, particularly regarding the qualitative aspects of such assessment. Despite this drawback, the Clement et al. study is a fine example of needed expressive test research conducted to explicitly replicate findings, in this case regarding differential diagnosis with older persons. In the Clement et al. work, attention to matching participants across diagnostic groups is also a strength.
Such work underscores the importance of expressive personality research with both intact
and cognitively/physically impaired older persons (e.g., those suffering from dementia or cognitive impairment) to more fully understand the extent of age-relatedness in expressive personality test performance. Indeed, Wang, Ericsson, Winblad and Fratiglioni (1998) observed that among persons with both cognitive impairment and severe dementia, mental, visual, and physical problems often compromise persons' ability to draw. Still, even if these other problems are considered, figure drawings can apparently still differentiate older persons with varying degrees of cognitive impairment. This is especially important when cognitive declines are mild, perhaps as a precursor to full-blown diagnosable dementia.
Discussion:
The Utility of Human Figure Drawing Techniques with Older Adults 30 Years On:
The primary effectiveness of drawing techniques seems to be in differentiating persons with dementia/Alzheimer's disease from normal (nonpatient) older adults, and in this respect, the above findings are quite impressive, and consistently robust across studies. We note that the vast majority of effect sizes (Cohen's d, r) reported here are of at least moderate strength (d = .50, r = .30). Although some samples are small, these often use matching designs and/or statistical controls of at least some extraneous variables, and some of the sample sizes are in the hundreds, quite large by clinical research standards. Moreover, the work reviewed here reflects a serious attempt at replication/cross validation of findings, based upon the studies by Clement et al., Silver, and Ericsson and his colleagues.
Still, stronger research designs are needed that consistently utilize larger, more representative samples, analyze and report cultural differences, and match groups on moderator variables (e.g., level of education, gender, health status, psychomotor skill, physical decline, sensory impairments, both premorbid
and disease-related drawing competence) that would otherwise confound group comparisons. Controls for drawing competence per se and a lack of confidence in one's drawing skills, as well as for age-related or disease-specific (e.g., arthritis) shifts in drawing ability remain relatively lacking in the HFD literature with older adults. This is an especially noteworthy limitation in that Ericsson et al. (1996) indeed found small but reliable age declines in the ability to copy geometric figures among non-demented persons aged 75-90+. Thus, the systematic impact (if any) of such factors on drawing techniques' sensitivity to detect dementia at the individual level remains to be established. Though older adults were not their focus, Gardner and Brown (2010) have considered possible methodological confounds in HFD research: the presence/absence of clothing or body details and whether stimulus figures being judged for participants' body shape and size preferences are presented in an orderly or disorderly fashion (and subject to spatial bias effects). These issues have yet to be explored in such research with older adults.
Given these common confounds, we were surprised that such techniques continue to be used as frequently as they are and that studies produced such large effect sizes. However, that HFD techniques do differentiate diagnostic groups (both in small scale and large scale studies) is a testament to the development and reliable application of scoring criteria that identify objective features which are not dependent upon drawing ability, artistic training, talent, self-confidence, and other extraneous individual differences.
Future research should flesh out the two areas we found most lacking compared with the previous reviews: age-normative data and psychometric information about the scoring systems used. As recent normative data for both normal and clinical samples of older adults are unavailable, this deficit needs to be rectified, and should include differentiation by
the moderator and control variables listed above. However, this is only important to the extent that HFD scores are used nomothetically
- in making comparisons of persons with regard to the normative tables of scores for their reference group. As with other expressive techniques (Panek et al., 2013), psychometric work is needed to develop a psychometric theory appropriate to these less structured measures in which the data are free-response behavior samples that are scored using a descriptive manual that quantifies certain features of the behavior sample. Psychometric standards, chiefly interscorer reliability, may then be applied to the results produced by the manual, rather than to the behavior samples. The movement in both the aging and personality literatures toward thinking about personality processes (Bornstein, 2007, 2009, 2011; Hooker & McAdams, 2003), as well as greater emphasis upon interindividual variability in personality (Mroczek, Spiro, & Griffin, 2006), may offer a promising new psychometric paradigm for evaluating HFD techniques.
Age Differences:
It is clear that HFD techniques are age-sensitive, in contrast to other expressive techniques (see Panek et al., 2013) which are not. Indeed, many age-related findings are consistent with a pattern of increasing personality constriction with increased age, perhaps in response to declines in physical health, cognitive capacity, everyday functionality, or lack of environmental stimulation. Interpretatively, we note that such findings may reflect one's reaction to cognitive and other impairments. Persons may be differentially sensitive to the age or disease-related loss of their decision-making and everyday functional skills and ultimately the loss of the self as an organized entity. Each of these changes is likely to lead to increased isolation from others and/or unwanted dependence upon them, and consequently predispose one to
depression and/or self-destructive thinking (Hayslip, Hicks-Patrick, &Panek, 2011; Miller, 1979).
In this respect, a significant deficit in the HFD older adult literature is a lack of longitudinal studies assessing age changes; we found but one exception to this void in this review (Jungwirth et al., 2009). As most studies to date are cross-sectional and often based upon small, non-representative samples, large-scale longitudinal studies are sorely needed to assess individual differences in intraindividual changes over time. Longitudinal work might inform us more clearly about whether personality does or does not change with age, depending in part upon the level (i.e. intrapsychic versus socioadaptational) at which personality is assessed (see Hayslip, et al., 2011; Mroczek, Spiro, & Griffen, 2006;
Neugarten, 1976, 1977).
This lack of longitudinal work is particularly unfortunate given that one easily implemented application of HFDs is to elicit HFDs from individuals at admission to a residential situation or facility as part of routine intake procedures. If change in functioning over time is suspected, new HFDs can then be compared with the intake baseline to evaluate change, much as a baseline mammogram is used to detect developing pathology. Such a strategy controls for many of the individual differences (e.g., premorbid drawing ability) that contribute to error variance in HFD scores. The specific features that differ between the first and second administration may help to differentiate economically among possible causes of change.
An additional direction in accurately assessing age effects in HFD performance is the differentiation of age effects from cohort effects (see Baltes, Reese, &Nesselroade, 1988), and thus, there is a lack of work exploring potential cohort/ historical influences on HFD performance. This is important given the
steadily rising average level of education in successive American cohorts. Prospective studies differentiating the personality process antecedents of cognitive or functional ability changes and/or terminal drop (nearness to death) from their consequences would also be quite valuable (see Mroczek et al).
It should be noted that in cases where the impact of age on HFD performance is of interest, the data must be considered correlational, wherein that age-related change is a dependent variable (see Baltes, Reese, & Nesselroade, 1988). Thus, dependent upon the acknowledgement and influence of moderating variables, findings regarding age differences may or may not replicate across studies. Consequently, the many studies that have examined age effects may only be partially consistent with those of Neugarten (1976, 1977) suggesting increased interiority with age at the intrapsychic level of personality functioning. Findings speaking to the possibility of cohort-related shifts in HFD performance could easily help to explain the failure to replicate earlier work.
Bearing on the replication of age differences, we also note the need for cross-cultural research investigating age differences in other cultures/countries utilizing HFD techniques, allowing us to evaluate the generalizability and/or cultural specificity of what we know about personality processes and aging in the
U.S. (Panek, 2006). This is important in light of the increasing attention to minority aging and to health disparities among minority older persons (Yee & Chiriboga, 2007). At the minimum, it is important to develop norms for ethnic groups within a particular country or culture (Panek, 2001).
Considering the question of differential diagnosis as well as developmental change, our review underscores the importance of HFD research with both intact and cognitively
/physically impaired older persons (e.g., those
suffering from dementia or mental retardation) to more fully understand the extent of age-relatedness in expressive personality test performance; intact and impaired persons are likely to age differently.
Comparing Diagnostic Groups:
The literature reviewed here clearly indicates that HFD techniques are being used by clinicians and researchers to assess older persons reliably. While the effectiveness and appropriateness of any measure in screening for dementia and/or evaluating persons' responses to aging-related change does vary by the purpose for which a specific test is utilized, the evidence shows that HFD techniques can indeed differentiate older persons with varying degrees of cognitive impairment. Importantly, such techniques may be valuable as well in uncovering the elder's response to the loss of cognitive skills. For these specific purposes, the weight of the evidence favors these constructive approaches over verbal expressive techniques (see Panek et al., 2013). Indeed, subtle personality shifts prior to the formal diagnosis of Alzheimer's disease which are characteristic of Mild Cognitive Impairment (Peterson et al., 2001; Shigeta& Homma, 2007) may be especially susceptible to detection via HFD techniques, as may one's reaction to normal age-associated memory impairments (see Crook, Bartus, et al., 1986). Such information may be predictive of adherence to treatment as well as treatment response among those who are adherent.
With a few exceptions, however, this work rests upon one-shot studies in dire need of replication, and in a few cases, the samples employed are quite small, underpowering findings. This remains true since it was observed by Kahill (1984) in her evaluation of HFD techniques (irrespective of age), based upon the literature from 1967-1982. Yet, we find here that there is substantial consistency in the question of differential diagnosis across studies utilizing samples of varying sizes using different
scoring approaches, and the effect sizes for many studies' findings are impressive enough to be taken seriously. The pattern of these findings across conceptual replications or near-replications suggests that they do indeed reflect real phenomena that may be of clinical importance to those working with older persons.
Given this body of evidence, we nevertheless argue that further HFD research is needed in the differential diagnosis of older adults, substantiating degrees of personality deterioration in such persons as well as among those whose physical health has worsened due to normal aging-related changes or due to pathological processes, i.e., illness. For example, as we have noted above, what may be operating in finding a decline in both cognitive and physical function is that we are assessing the older person's reaction to the loss of his/her skills. This contrasts with an older view of constructive test performance as a direct measure of cognitive functioning, a conclusion reached with regard to the use of HFD with children (reference). Importantly, the literature lacks studies in which HFD techniques and self-report inventories are jointly administered to clinical and non-clinical samples of older adults. This would enable one to triangulate the strengths and weaknesses for each test (Ganellen, 1996; Meyer, 1997).
Future Research Needed: Extraneous Individual Differences:
We found no studies focused on non-age-related, non-diagnostic individual differences since the previous reviews. In light of the potential confounding of age effects and extraneous influences on HFD performance, there is a continued need for research to determine the effects of extraneous factors (as mediator or moderator variables) such as sensorimotor impairment, health conditions, and psychosocial problems (e.g., isolation, loneliness, discrimination) on the HFD protocols of older adults. Generally speaking, some work with HFDs reflects these ends,
important in that such factors can be impediments to the use of such techniques in both research and clinical contexts (Anastasi & Urbina, 1997). However, in light of the current review, we advocate their exploration as important sources of useful information about interindividual variability among older adults (Nelson & Dannefer, 1992) that might require intervention or predict everyday functionality, and thus, would be important guides and caveats for treatment planning. Moreover, by controlling for the effects of such variables, triangulating by the use of multi-modal assessment, and/or observing their impact on expressive test responses, age differences/age changes in characteristic responses to such expressive techniques (see above) can be accurately assessed. Since the mid-1980s, researchers have moved away from simply investigating “typical” or “characteristic” response patterns of older adults on many verbal expressive personality tests (see Panek et al., 2013) to evaluating the effects of potential moderator variables on the responses of older adults.
An issue not unique to expressive assessment but a risk in the assessment of older persons that might be overlooked in clinical training is the possible misinterpretation of findings. Some of the moderator variables discussed above are far more common among older adults than in the general population, and failure to differentiate their effects from age-related change, cognitive debilitation, or emotional functioning in older persons may result in misdiagnosis, inappropriate treatment, or a failure to treat. Future research studies should either build such variables into their designs, control for them as confounds, or consider them in the interpretation, as appropriate to the sample and the study's purpose.
Future Research Needed: Normative Data:
Although dated norms for normal and clinical samples of older adults exist for human figure drawing techniques (see Hayslip & Lowman,
1986; Panek et al., 1983), to the extent that normative comparisons are desired, new norms for current cohorts are needed. Indeed, norms may be unnecessary when interpretation is process-oriented, content-based, ipsative, or idiographic.
Concluding Remarks and Recommendations:
Given the difficulties in assessing impaired older persons with self-report methods, and the strength of the findings gathered here, we advocate the use of HFD techniques in tapping personality processes, i.e. transient states, self-regulation, that covary with situational/ contextual demands (see Hooker & McAdams, 2003), thus marrying non-trait-like views of personality to HFD research. As HFD scores may represent older persons' emotional responses to cognitive decline/impairment, such work would be quite valuable in assessing the functioning of persons with dementia who cannot or will not submit to being assessed more traditionally or to being extensively interviewed. Thus, in our opinion HFD techniques, have unique value for older adults, whether used alone or as an element of multimodal assessment (see Edelstein & Segal, 2011).
Good assessment data are valuable when decisions affecting older adults and their families need to be made regarding a) the wisdom of admission to a long-term care facility,
b) the level of care one might need, c) one's capacity to cope with and/or represent physical or cognitive deficits, or d) the implementation and evaluation of treatment plans and interventions. HFD techniques may play a valuable role in these aspects of patient care and allow the clinician to generate a hypothesis regarding the patient's functioning which can be investigated, monitored and modified with additional data as required. The course of an older person's emotional reactions to a loss of intellect or memory over time, for which HFD techniques might be quite important (e.g. see
the above discussed work by Silver and Ericsson and his colleagues, Morin & Benshalah, 1998). Such changes reflects altered views of the self, dimishedoptimisim about life, and changes in one's relationships to others; these are likely to be of central interest to the clinician, as might aspects of one's physical identity/body image representing how one appears to self and others as well as reflecting one's physical competence. It ihas been pointed out that how older persons cope with such challenges also has important implications for not only their own health and well-being, but also for that of their caregivers (Knight, Kaskie, Shurgot, & Dave, 2006; Knight & Losada, 2011).
Although the technology of manualized scoring systems and training manual development is less mature for HFDs than for other expressive techniques (e.g., Rorschach comprehensive system, Exner, 2001; thematic apperceptive techniques, Jenkins, 2008a, 2008b), the need for systematic, objective scoring procedures for HFDs was recognized half a century ago, and some early manuals for the general population are still in use (Goodenough, 1962; Jones & Rich, 1957; Hammer, 1997; Machover, 1949). The forthcoming volume edited by Handler and Thomas (2014) includes chapters that present scoring manuals by Naglieri and by Crisi, among others, that should be studied with older adult samples. However, many manuals, especially the earlier ones, are dependent for accurate scoring on the interpreter's clinical experience and/or grasp of theory that give different clinicians a similar set of working assumptions for scoring. In the present clinical environment of theoretical diversity (indeed differential emphasis on theory) and demand for convergent empirical evidence, training materials and methods must become more efficient and less subject to divergent interpretation. Explicit manuals with many concrete examples and rationales for scoring are needed to make accurate use of these techniques accessible to students, practicing
clinicians, and researchers, as well as defensible to critics.
A helpful clinical strategy while we wait for the necessary science, and a good strategy for producing that science, is the joint use of tests and assessment procedures that are sensitive to the same diagnostic targets but are susceptible to different sources of systematic method error variance. Apparently “conflicting” findings among methods can reveal a need for the careful differential diagnosis of age-normative, pathological, and situational influences. The best example of this strategy, i.e., reviewing the combined use of the Rorschach and MMPI-2, given their different response style biases, is work reviewed by Ganellen (1996) (see also Clemence& Handler, 2001). An important emerging strategy is an increasing focus on the elements of the response process, for example their sequencing or overall integration, relative to a single quantitative score (Bornstein, 2011). Such integration has long been an important implicit element of most of the extant HFD scoring systems; it remains only to make the scoring criteria explicit and the manuals and training materials widely available.
As 1) the above cited work could easily be seen as an underestimate of the true worth of HFD methods with older adults, 2) such persons are increasingly more likely to be seen in clinical settings and 3) the need to assess them becomes more imperative, further appreciation of the unique value of HFD techniques with older persons will hopefully increase. Their ease and efficiency of administration, engaging and less demanding assessment process, non-threatening indirectness, ambiguity to the client, and subtle data yielded recommend them for this population. HFDs indeed merit much more research attention regarding the study of personality and aging.
References:
Abeles, N. (1997). What practitioners should know about
working with older adults.APA. Washington, DC. Anastasi, A., & Urbina, S. (1997). Psychological testing (7th
Ed.). Upper Saddle River, NJ: Prentice Hall.
APA Monitor. Psychology's growth careers, April 2008, pp.
64-65.
Baltes, P., Reese, H., &Nesselroade, J. (1988). Life span developmental psychology: Introduction to research methods. Hillsdale, NJ: Earlbaum.
Blazer, D. (2002).Depression in late life.New York: Springer.
Bornstein, R.F. (2007). Toward a process-based framework for classifying personality tests: Comment on Meyer and Kurtz (2006). Journal of Personality Assessment, 89, 202-207.
Bornstein, R. F. (2009). Heisenberg, Kandinsky, and the heteromethod convergence problem:Lessons from within and beyond psychology. Journal of Personality Assessment, 91, 1-8.
Bornstein, R. F. (2011). Toward a process-focused model of test score validity: Improving psychological assessment in science and practice. Psychological Assessment, 23, 532-544.
Clemence, A. J., & Handler, L. (2001). Psychological assessment on internship: A survey o f t r a i n i n g directors and their expectations for students. Journal of Personality Assessment, 76, 18-47.
Clément, J.-P., Marchan, F., Boyon, D., Monti, P., Léger, J.-M., & Derouesné, C. (1996). Utilization of the Draw a Person Test in the elderly. International Psychogeriatrics, 8, 349-364.
Crook, T.H., Bartus, R.T. et al. (1986). Age-associated memory impairment: Proposed diagnostic criteria and measures of clinical change. Developmental Neuropsychology, 2.261-276.
D o u g h e r t y, L . M . , & C o o k , D . B . ( 1 9 9 1 , November).Development of a geriatric draw-a-person scoring system. Paper presented at the Annual Scientific Meeting of the Gerontological Society of America. San Francisco, CA.
Ericsson, K., Forssell, K., Amberla, K., Holmén, K., Viitanen, M., &Winblad, B. (1991). Graphic skills used as an instrument for detecting higher cortical dysfunctions in old age.Human Movement Science, 10, 335-349.Ericsson, K., Hilleräs, P., Holmén, K., Jorm, A., et al. (1994).The short human figure drawing scale for evaluation of suspect cognitive dysfunction in old age. Archives of Gerontology and Geriatrics, 19, 243-251.
Ericsson, K., Hilleräs, P., Holmén, K., & Winblad, B. (1996).Human-figure drawing (HFD) in screening of cognitive impairment in old age. Journal of Medical
Screening, 3, 105-109.
Ericsson, K., Hilleräs, P., Sundell, M.-L., & Winblad, B. (1997). Human-figure drawings from 4 to 104 and in people with impaired cognition. International Journal of Practical Approaches to Disability, 21, 8-15.
Ericsson, K., Winblad, B., & Nilsson, L. (2001). Human figure-drawing and memory functioning across the adult life span. Archives of Gerontology and Geriatrics, 32, 151-166.
Exner, J. E., Jr. (2001).A Rorschach workbook for the comprehensive system.Fifth ed. Asheville, NC: Rorschach Workshops.
Fan, N., He, H., Xiao, J., & Jia, Y. (2002).Cognitive impairment in patients with Alzheimer's disease. Chinese Mental Health Journal, 594, 590-591.
Finn, S. E. (2007). In our clients' shoes: Theory and techniques of therapeutic assessment. Mahwah, N.J: Lawrence Erlbaum.
Fischer, C. T. (1994). Individualizing psychological assessment. Hillsdale, NJ: Lawrence Erlbaum.
Freeman, R.Q., Giovannetti, T., Lamar, M., Cloud, B., Stern, R., Kaplan, E., & Libon, D. (2000). Visuoconstructive problems in dementia: Contribution of executive systems functions. Neuropsychology, 14, 415-426.
Ganellen, R. J. (1996). Integrating the Rorschach and the MMPI-2 in personality assessment. Mahwah, NJ: Lawrence Erlbaum Associates.
Gardner, R.M., & Brown, D. L. (2010). Body image assessment: A review of figural drawing scales. Personality and Individual Differences, 48, 107-111.
Goodenough, F. L. (1962). The measurement of intelligence by drawings. New York: World Book.
Hammer, E. F. (1997). Advances in projective drawings interpretation. Springfield, IL: Charles C. Thomas.
Handler, L., & Thomas, A. D. (Eds.) (2014). Drawings in assessment and psychotherapy: Research and application. New York, NY: Routledge.
Hayslip, B., Jr. (1994). The Hand Test and aging. In E. E. Wagner & G. R Young (Eds.), The Hand Test: A casebook for advanced interpretation (pp.167-182). Los Angeles, CA: Western Psychological Series.
Hayslip, B., Jr., Cooper, C. C., Dougherty, L. M., & Cook, D.
B. (1997). Body image in adulthood: A projective approach. Journal of Personality Assessment, 68, 628-649.
Hayslip, B., Jr., & Lowman, R. L. (1986). The clinical use of projective techniques with the aged: A critical review and synthesis. Clinical Gerontologist, 5, 63-94.
Hayslip, B., Jr., Panek, P. E., & Hicks-Patrick, J. (2011).
Adult development and aging (5th Ed.). Malabar, FL: Kreiger Publishing Company.
Hayslip, B. J., Jr., Panek, P. E., & Stoner, S. B. (1990). Cohort differences in Hand Test performance: A time lagged analysis. Journal of Personality Assessment, 54, 704-710.
Helmes, E. (2012). Cognitive screening of older adults: the utility of pentagon drawing. International Psychogeriatrics, 25, 413-419.
Holtzman, W.H., Thorpe, J.S., Swartz, J.D., & Herron, E.W. (1961). Inkblot perception and p e r s o n a l i t y : Holtzman Inkblot Technique. Austin, TX: University of Texas Press.
Hooker, K., & McAdams, D. (2003). Personality reconsidered: A new agenda for aging research. Journal of Gerontology: Psychological Sciences, 58B, P296-P304.
Jenkins, S. R. (2008a). Teaching how to learn reliable scoring. In S. R. Jenkins (ed.), Handbook of clinical scoring systems for thematic apperceptive techniques (pp. 39-66). Mahwah, NJ: Lawrence Erlbaum.
Jenkins, S. R. (2008b). Handbook of clinical scoring systems for thematic apperceptive techniques. Mahwah, NJ: Lawrence Erlbaum.
Jones, A. W., & Rich, T. A. (1957). The Goodenough Draw-A-Man Test as a measure of intelligence in aged adults. Journal of Consulting Psychology, 1, 235–238.
Jungwirth, S., Zehetmayer, S., et al. (2009). Screening for Alzheimer's dementia at age 78 with short psychometric instruments. International Psychogeriatrics, 21, 548-
558.
Kane, R. L., & Kane, R. A. (2000). Assessing older persons: Meaning and practical applications. New York: Oxford.
Knapp, N. M. (1994). Research with diagnostic drawings for normal and Alzheimer's subjects. Art Therapy: Journal of the American Art Therapy Association, 11, 131-138.
Knight, B. (2004).Psychotherapy with older adults (3rd Ed.).
Thousand Oaks, CA: Sage.
Knight, B., Kaskie, B., Shurgot, G., & Dave, J. (2006).Improving the mental health of older adults.In J. E. Birren& K. W. Schaie (Eds.), Handbook of the psychology of aging (6th Ed.) (pp. 408-425). San Diego: Academic Press.
Knight, B., & Losada, A. (2011). Family caregiving for cognitively and physically frail older adults: Theory, research, and practice. In K.W. Schaie& S.L. Willis (Eds.), Handbook of the psychology of aging: Seventh edition (pp. 353-366). San Diego, CA: Academic Press.
Koenig, M., George, L., & Schneider, R. (1994). Mental
health care for older adults in the year 2020: A dangerous and avoided topic. The Gerontologist, 34, 674-679.
Lilienfeld, S. O., Wood, J. M., & Garb, H. N. (2000).The scientific status of projective techniques. Psychological Science in the Public Interest, 1, 27-66.
Machover, K. (1949). Personality projection in the drawing of the human figure. Springfield, Ill: C.C. Thomas.
Meyer, G. J. (1997). On the integration of personality assessment methods: The Rorschach and MMPI. Journal of Personality Assessment, 68, 297-330.
Miller, M. (1979). Suicide after 60: The final alternative. New York: Springer.
Moore, V., & Wyke, M.A. (1984).Drawing disability in patients with senile dementia. Psychological Medicine, 14, 97-105.
Mroczek, D., Spiro, A., & Griffen, P. (2006).Personality and aging, In J. Birren& K. W. Schaie (eds.), Handbook of the psychology of aging (6th Ed.) (pp.363-378). San Diego, CA: Academic Press.
Nelson, E., & Dannefer, D. (1992). Aged heterogeneity: Fact or fiction? The fate of diversity in gerontological research. The Gerontologist, 32, 17-23.
Neugarten, B. L. (1976). Adaptation and the life cycle.
Counseling Psychologist, 6, 16-20.
Neurgarten, B.L. (1977). Personality and aging. In J. Birren & K. Schaie (eds.), Handbook of the psychology of aging (pp. 626-649). New York: Van Nostrand Reinhold.
Panek, P. E. (2006). The aging population: A challenge for projective psychology (Editorial). Journal of Projective Psychology & Mental Health, 13, 5-6.
Panek, P. E., Jenkins, S.R., Hayslip, B., &Moske, A. (In Press). Expressive personality testing with older adults: 25 years later. Journal of Personality Assessment.
Panek, P. E., Wagner, E. E., & Kennedy-Zwergel, K. (1983).A review of projective test findings with older adults. Journal of Personality Assessment, 47, 562-582.
Peat, C. M., Peyrel, N.L., & Muehlenkamp, J. J. (2008). Body image and eating disorders in older adults: A review. The Journal of General Psychology, 135, 135-148.
Peterson, R. C., Doodey, R., Kurz, A., Mohs, R., Morris, J., & Rabin, P. (2001). Current concepts in mild cognitive impairment. Archives of Neurology, 58, 1985-1992.
Pruis, T.A., & Janowsky, J.S. (2010).Assessment of body image in younger and older women. The Journal of General Psychology, 137, 225-238.
Reichlin, R. E. (1984). Current perspectives on Rorschach performance among older adults. Journal of Personality Assessment, 48, 71-81.
Roback, H. B. (1968). Human figure drawings: Their utility in the clinical psychologist's armamentarium for personality assessment. Psychological Bulletin, 70, 1-
19.
Scogin, F. (2007).Introduction to the special section on evidence-based psychological treatments for older adults. Psychology and Aging, 22, 1-3.
Shigeta, M., & Homma, A. (2007). Alzheimer's disease. In J. Blackburn, C. Domas (eds.), Handbook of gerontology: Evidence-based approaches to theory, practice, and policy (pp. 333-336). New York: Wiley.
Silver, R. (1986).Stimulus drawings and techniques. New York: Trillum Press.
Silver, R. (1987). Sex differences in the emotional content of drawings. Art Therapy, 4, 67-77.
Silver, R. (1991). Stimulus drawings & techniques.
Sarasota. FL: Albin Press Distributors.
Silver, R. (1993a). Age and gender differences expressed through drawings: A study of attitudes toward self and others. Art Therapy: Journal of the American Art Therapy Association, 10, 159-168.
Silver, R. (1993b). Assessing the emotional contents of drawings by older adults: Research findings and implications. American Journal of Art Therapy, 32, 46-52.
Starr, B. D., & Weiner, M. B. (1993). The Projective Assessment of Aging Method (PAAM) in clinical practice. Clinical Gerontologist, 12, 69-80.
Swenson, C. H. (1957). Empirical evaluations of human figure drawings. Psychological Bulletin, 54, 431-466.
Wang H.-X., Ericsson, K., Winblad, B., & Fratiglioni, L. (1998). The Human Figure Drawing test as a screen for dementia in the elderly: A community-based study. Archives of Gerontology and Geriatrics, 27, 25-34.
Wells, C. (1979). Pseudodementia. American Journal of Psychiatry, 136, 895-900.
Wolber, G. (1980). A practical approach to the psychological evaluation of elderly patients. Perceptual and Motor Skills, 51, 499-505.
Yee, B., & Chiriboga, D. A. (2007).Issues of diversity in health psychology and aging. In C. M. Aldwin, C. L. Park, & A. Spiro III (eds.), Handbook of health psychological aging. (pp. 286-312). New York: Guilford.
Yon, A., & Scogin, F. (2007). Procedures for identifying evidenced-based psychological treatments for older