This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation

Do Middle School Students Really Have Fixed Images of Elders?

¹ General Clinical Research Center
² Sam and Ann Barshop Institute for Longevity and Aging Studies
Divisions of ³ Geriatrics and Gerontology
⁴ Nephrology, Department of Medicine, University of Texas Health Science Center at San Antonio.

Address correspondence to Michael Lichtenstein, Department of Medicine, General Clinical Research Center, University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Dr., San Antonio, TX 78229-3900. E-mail: lichtenstei{at}uthscsa.edu

	Abstract

TOP Abstract Human Figure Drawing and... Methods Results Discussion References

 
Objectives. The purpose of this study was to determine whether combinations of characteristics, abstracted from drawings of elders made by middle school students, grouped together to form cohesive perceptions, or stereotypes, of human aging.

Methods. We abstracted 49 characteristics from drawings of elders made by 1,944 students at two middle schools in San Antonio, Texas, at the beginning of the 1998–1999 school year. Correlational and factor analyses were used to determine if there was an underlying structure or grouping to the characteristics. Logistic regression was used to determine the variables associated with the investigators' classification of the images as positive, neutral, or negative.

Results. The standardized alpha coefficient for the 49 variables was low ( = 0.37). The Spearman rho correlations between the variables were also low, with 90.2% of the 1,176 comparisons being < 0.10. Exploratory factor analyses did not provide a useful grouping of characteristics drawn by the students, including analyses stratified by gender and restricted to the most common 34 characteristics. Among the 49 characteristics that emerged from the drawings, 11, 4, and 11 traits were directly associated with classifying the drawings as positive, neutral, or negative, respectively.

Discussion. These analyses indicate that middle school students have not formed strong images regarding aging: No clear cohesive stereotypes of elders emerged from the images drawn by these children. Absence of stereotypic views implies that middle school students may not have a built-in bias toward older people and age-associated changes. This suggests that young adolescents are at a point where instruction including gerontological content can be used to effectively teach about aging and health promotion.

STEREOTYPES are fixed, simplified characterizations of groups of humans. Walter Lippmann (1922) first described stereotypes as devices that simplify and provide order to a complex world. The presumption is that knowledge of a particular trait (e.g., advanced chronologic age) allows attribution of other characteristics to an individual within that group. People quickly develop coherent impressions about another person's attitudes and behavior based on a few salient cues (Grant, Ross, Button, Hannah, & Hoskins, 2001). In so doing, people develop a sense of security that they "know" something based on a limited set of facts or experience. Lippmann (1922) asserted that patterns of stereotypes are seldom neutral and that their existence is a barrier to education.

Throughout the literature exploring age-related perceptions, various terms have been used interchangeably in reporting "stereotypes." These include impressions (Grant et al., 2001), perceptions (McTavish, 1982; Mitchell, Wilson, Revicki, & Parker, 1985), attributions (Bassili & Reil, 1981) or attributes (Grant, Button, Hannah, & Ross, 2002), traits (Bruner, Shapiro, & Tagiuri, 1958; Field & Gueldner, 2001) or trait descriptors (Perdue & Gurtman, 1990), prototypic features (Brewer, Dull, & Lui, 1981), attitudes (McTavish, 1982) or impressions of attitudes (Grant et al., 2001), beliefs (McTavish, 1982), common misconceptions (McTavish, 1982) or myths (Field & Gueldner, 2001; Powell, 1998), and orientations (McTavish, 1982). Although operational definitions for these terms exist within each publication, they typically are not used consistently across studies, making comparisons between bodies of work difficult.

The term "person perception" was coined by social researchers Tagiuri and Petrullo in 1958 to describe aspects of the ways in which humans perceive and make judgments about others. They proposed that the perceiver infers properties and potentialities of a person through observations made about "intentions, attitudes, emotions, ideas, abilities, purposes, traits—events that are ... inside the person" (p. x). Though perceivers make observations of other people's actions, the perceiver's own presence and behavior alter the perceptual characteristics of the person whose state is being judged. This creates a double interaction that is different from observing an inanimate object. In making judgments about other people, perceivers use their own experiences, real or vicarious, in forming those opinions. Given the challenges in assessing a person's view of others, it is incumbent on researchers to use means of soliciting free and unrestricted descriptions of other persons (Tagiuri & Petrullo, 1958). Free and unrestricted descriptions may be obtained with the experimenter assigning the "class" of the person to be described, but providing no other parameters. This potentially avoids problems of framing expectations ahead of time with trait lists or anchored Likert Scales.

By seeking to determine the characteristics that persons associate with older adults, Tuckman and Lorge (1953) began the efforts to describe an old-age stereotype. Over time, teams of investigators have reported a variety of old-age stereotypes. Palmore (1998, 1999) has described "ageism" as taking two broad forms: positive and negative. Descriptions of older persons vary in their negative tone, ranging from inflexible or withdrawn, passive and dependent, nurturing or religious, to displaying forms of physical and mental deterioration (Bassili & Reil, 1981). Some descriptions are even more negative, reporting the old-age stereotype as persons who are senile with memory loss, rigid personalities, depressed, lonely, isolated, and institutionalized by their families (Field & Gueldner, 2001). In contrast, on the positive side, data from the Berkeley Older Generation Study suggest that older adults are heterogeneous, having potential for healthy lives in spite of physical disability and showing little decline in enduring personality traits (Field & Gueldner, 2001). Brewer and colleagues (1981) described commonly held traits of elders as inconsistent (irritable/serene, suspicious/naively trusting, conservative/eccentric) and not yielding a single prototype. Using photographs and descriptive statements in their study with college-age students, they determined three types of elders: grandmotherly, elder statesman, and senior citizen (Brewer et al., 1981). A 1990 study (Perdue & Gurtman) delineated several negative old-age stereotypes but noted that negative representations are not universal. Hickey, Hickey, and Kalish (1968) reported that ageism in children and adolescents revealed a mixture of positive and negative attributes in spontaneous descriptions of elders. They also suggest that ageism occurs only when individuals are induced to judge elders as compared with the young. Perdue and Gurtman (1990) examined the idea that the setting in which social information is encoded may result in automatic or unconscious biases about age. The possibility exists that the strength of the old-age stereotypes among elderly persons reflects a sense of inevitability (Bassili & Reil, 1981; Hausdorff, Levy, & Wei, 1999). However, it may be that older adults use negative age stereotypes as more of a standard of comparison rather than actually integrating the images into their own self-perceptions (Pinquart, 2002).

MacTavish's (1971) extensive review of perceptions of old people documented the heterogeneity of prior investigations in this area. He discussed society-level and individual-level studies and stated that many of the negative views of elderly people in the United States may have been overemphasized by the gerontologists of the time. McTavish recommended that researchers should (a) draw from larger and more carefully drawn samples of subjects from a defined sampling frame, (b) use more consistent and technically examined research instruments, and (c) use conceptually integrated proposals for linking cause, correlates, and outcomes.

In 1982, McTavish reviewed the measures of perception about old people used since 1942. He organized the instruments used into four categories: Yes–No Scales, Likert-Type Agree–Disagree Scales, Semantic Differential Scales, and Sentence-Completion and Content Analytic Procedures. In the 1970s and 1980s, researchers introduced the use of photographs, line drawings, and artists' renditions of photographs to elicit stereotype information. Investigators (Di Leo, 1983; Koppitz, 1968) obtained drawings from disabled children to learn to about their perceptions of family and self. In growing older, the child's interests and feelings extend to those outside the family circle, but family continues to be the most important influence (Di Leo, 1983). In the 1990s, investigators began soliciting drawings from participants with differing instructions for the construction of those drawings, with each looking for images of "old." The categories of tools used by a sample of research groups since McTavish's review are summarized in Table 1. The table points out the variability in assessment techniques and study subjects that persists, underscoring the difficulties in comparing findings across studies.

	HUMAN FIGURE DRAWING AND ASSESSMENT OF CHILDREN'S ATTITUDES ABOUT AGING

TOP Abstract Human Figure Drawing and... Methods Results Discussion References

 
There is a rich history of the use of children's human figure drawing from the late 1800s until the present time. The first reproductions of children's art appeared in the writings of Corrado Ricci, a renowned art critic, in 1887. Ricci is credited with having insight into the psychological implications of human figure drawings by children (Klepsch & Logie, 1982).

In the early 20^th century, Goodenough, a child psychologist and faculty member at the University of Minnesota, developed the "Draw-a-Man Test" (Goodenough, 1926). Goodenough believed it was possible to estimate the intelligence of young children based on their drawings of human figures. The test, which produced a single-score measure of mental ability, was used to survey the intellectual status of young children and to study those with hearing handicaps and suspected neurologic deficiencies. Over time, it was used to study adjustment and personality problems, intellectual development, attitudes, delinquency, and character defects (Di Leo, 1983; Harris, 1963; Harris & Pinder, 1974). Through drawing, children share their internal world of experiences. Their drawings reflect their interpersonal development and how they respond to the world around them—their relationships with family members, friends, and those within the larger community (Malchiodi, 1998).

Artists' and children's drawings have been used in at least four studies to explore children's attitudes about elders (Couper, Donoforio, & Goyer, 1995; Falchikov, 1990; Mitchell et al., 1985; Seefeldt, Jantz, Galper, & Serock, 1977). Seefeldt and colleagues (1977) had 180 preschool to sixth grade children evaluate four drawings of one man at ages 20, 40, 60, and 80 years through a structured interview. The majority of the children (69%) correctly sequenced the drawings, thereby demonstrating an understanding of relative age. Interviews with the children indicated negative attitudes toward the physical characteristics of the image of the eldest man and the children's view of their own aging. Mitchell and colleagues (1985) presented three drawings of persons at different ages to 225 elementary school children. The children responded to a 25-item questionnaire designed to assess their perceptions about older adults. The responses were categorized by factor analysis into personality traits, affective relations, and physical abilities. Whereas there were differences in perceptions based on the age and gender of the drawn person (e.g., women less messy than men), the gender, race, and grade level of the children responding to the images did not affect the responses.

Rather than interpret presented drawings, Falchikov (1990) asked 28 Scottish children (mean age 11 years) to draw images of an old man, an old woman, a young man, and a young woman. When the drawings were compared, images of elders frequently showed signs of physical deterioration (e.g., the use of a cane, inactivity), and the older figures were drawn smaller than the younger figures. Falchikov interpreted the images as frequently portraying stereotypes of aging. This small study illustrated the rich detail of these drawings and contrasts the extremes of youth and old age. It provides a starting framework for other investigators working to describe the content of children's drawings of elders.

Couper and colleagues (1995) had 423 children (aged 6–11 years) draw pictures of an old person and a young person and then asked them to explain the differences. Children who drew images of someone they knew (e.g., a grandparent) were more likely to portray an older person with positive attributes. Girls tended to draw more positive images of elders than boys, whereas boys expressed more negative ideas about aging. On average, older children were more negative about old age than younger children.

The purpose of the current article is to determine whether combinations of traits, as they emerged and were abstracted from middle school students' drawings of elders, group consistently and meaningfully together to form an age-related stereotype. We chose to work with middle school students because early adolescence is characterized by a realignment of relationships between parents and children, with children striving to attain more autonomy and control (Holmbeck, Paikoff, & Brooks-Gunn, 1995). Adolescents are receptive to information about themselves and their bodies and anxious to become more independent in their decision making (Millstein, Petersen, & Nightengale, 1993). Thus, these children are at a point in their lives where health promotion activities have the potential for maintaining positive growth, healthy behaviors, and avoidance of risky behaviors (Parfenoff & Paikoff, 1997; Tanaka, 1996). Prior work by our group (Lichtenstein et al., 2003) demonstrated that middle school children are more likely to view their own futures more positively than those of their parents and other elders. However, middle school children infrequently associated specific diseases or conditions as markers of old age; that is, there was little evidence that these adolescents had knowledge about how illness may affect their health and well-being in the future. Believing it important to use a life course perspective for this work at a time when children are individuating within their families (Jackson & Sellers, 1997), we undertook to determine if there were specific pervasive stereotypes of elders that could be potential barriers to children learning about health promotion and disease prevention.

We suggest that stereotypes may be analyzed like syndromes in medicine, where a group of clinical features co-exist to a degree greater than would be expected by chance alone. To the extent that stereotypes may exist, the distribution and correlations among characteristics can be analyzed to determine if any are statistically associated with each other in a meaningful way. We reasoned that if stereotypes existed, certain characteristics would cluster together in a factor structure. If no underlying factor structure existed, this would provide evidence that, a priori, groups of children do not have strong established stereotypes about elders. In previous work (Lichtenstein et al., 2001), we classified middle school children's drawings of elders as positive, neutral, or negative. If the traits abstracted from the drawings did not cluster together in an exploratory analysis, we decided to analyze the underlying characteristics associated with labeling drawings as positive, neutral, or negative. This approach provided a method of classifying the children's emergent characteristics into the researchers' categories.

	METHODS

TOP Abstract Human Figure Drawing and... Methods Results Discussion References

 
Description of Schools
Northside Independent School District (NISD) is the sixth largest school district (of 1,110) in Texas, serving > 61,000 students in 350 square miles of Northwest San Antonio and surrounding Bexar County. The principals from the 12 NISD middle schools were surveyed about the schools and their willingness to participate in the project. Two schools, with distinctly different student populations, agreed to participate. Because prior analyses showed no difference in distribution of overall assessments of the drawings at the beginning of the school year (Lichtenstein et al., 2001), we pooled the drawing characteristics from both schools for this report. Of the 2,476 students enrolled in both schools (1,081 in School 1 and 1,395 at School 2) at the beginning of the school year, 50.7% were girls, 44.1% were Mexican Americans (63.4% in School 1, 29.1% in School 2), 48.2% were European American (28.5% in School 1, 63.5% in School 2), and 4.6% were African American (5.3% in School 1 and 4.0% in School 2). A greater proportion of children were economically disadvantaged (defined as being eligible for a federally subsidized school lunch program) at School 1 (43.6%) compared with School 2 (10.4%). The student/teacher ratios were 14:1 at School 1 and 16:1 at School 2.

Children's Drawings of Older Persons
The Positively Aging teaching materials use examples from geriatrics and gerontology to facilitate interdisciplinary learning at the middle school level (Lichtenstein et al., 1999). Unit 2 ("A Look at Them") is designed to explore characteristics that children associate with aging through an activity titled "Help the NIA." Students are first presented with background material about the mission of the National Institute on Aging (NIA). Next, they are told that the NIA has a problem: It does not know what a typical older person looks like. The teachers asked their students to help the NIA by drawing a picture of a typical older person. Teachers instructed their students to draw the whole person in a setting and to make the very best picture possible.

The students produced their drawings on a worksheet containing a 14.0 x 17.5–cm rectangle. Examples of student drawings illustrating positive, neutral, and negative images of elders are shown in Figures 1–3 (the drawings selected for this article are more detailed and carefully drawn than many of the images). After completing their drawings, students wrote written responses to a series of questions that further described the picture. The questions included the person's age, activities, feelings, thoughts, possible relation to the student, and how the drawn person's characteristics differ from those of the student.

View larger version (90K): [in this window] [in a new window]   Figure 1. Positive drawing of an older woman. The middle school student who drew the picture stated the woman was 63 years old and her grandmother. The description of the drawing stated, "This person is at a gym working out. She feels good and happy because she is keeping healthy. She teaches me all about cooking and family history."

View larger version (68K): [in this window] [in a new window]   Figure 2. Neutral drawing of an older man. The middle school student who drew the picture stated the man was 68–70 years old. The student did not draw someone he or she knew, and the only comment was, "He is watching TV at his home."

View larger version (96K): [in this window] [in a new window]   Figure 3. Negative drawing of an older man. The middle school student who drew the picture stated the man was 64 years old. The student did not draw someone he or she knew, and the only comment was, "He is mad, and he is yelling at someone."

Abstracting Characteristics From the Drawings
Our goal was to be as inclusive as possible and abstract as much information from each drawing as possible. Using a structured coding sheet and standardized instructions, four raters coded the characteristics of the drawings in detail. Forty-nine features were abstracted from the drawings and the accompanying descriptions written by the students. We coded the following characteristics of the drawn figures (Table 2): age, figure height (measured to 0.1 cm), gender, physical features (e.g., wrinkles), clothing style, facial expression, emotion, personality, loneliness, cognitive features (e.g., wisdom), health problems (e.g., weakness or specific diseases), use of physical aids, setting of the drawing, whether or not the figure was pictured with others, the figure's position (e.g., sitting), recreational or vocational activities (e.g., gardening or driving), activity level, and living arrangements. Most variables were coded as dichotomous (present or absent) with a few (activity level, age, position, setting, and height) being coded into ordinal variables. If a characteristic was not explicitly present or described, it was coded as not mentioned or absent.

Statistical Methods
We used a principal components exploratory factor analysis to determine if there was an underlying factor structure for the abstracted characteristics (Jackson, 1991; Kim & Mueller, 1978; Pett, Lackey, & Sullivan, 2003). Spearman rho nonparametric correlations were used to assess associations between the characteristics of the drawings (Altman, 1999). Coefficient alpha was used to assess the extent of internal consistency among the characteristics (Cortina, 1993; Cronbach, 1951).

To determine which characteristics were associated with the overall positive, neutral, and negative ratings of the drawings, we conducted three step-forward logistic regressions (Hosmer & Lemshow, 1989). First, the characteristics associated with a [positive rating] versus [neutral or negative ratings] were examined. The second regression examined [neutral ratings] versus [positive and negative ratings]. The third regression examined [negative ratings] versus [neutral and positive ratings]. To further reduce the number of variables, we selected characteristics that only increased the likelihood of assigning a rating to a drawing and reran a standard logistic regression (all associated variables entered the model together).

	RESULTS

TOP Abstract Human Figure Drawing and... Methods Results Discussion References

 
Response Rates
Drawing were obtained from 1,997 (80.6%) of the students at the beginning of the school year (82.5% at School 1, 79.2% at School 2). Within each school, there were no differences between responding students and nonresponders with respect to grade, ethnic group, gender, or being economically disadvantaged. Of the 1,997 drawings, 53 (2.7%) were missing ethnic and gender data for the students and were excluded. In another 34 drawings (1.7%), the gender of the figure could not be determined by inspection or interpretation of the description; these were retained in the tabulation of characteristics but excluded from the analyses. The correlational, factor structure, and regression analyses were based on the 1,910 drawings (77.2% of students, 95.7% of collected drawings) with complete coding and information.

Characteristics of Drawings
The categorization and frequency distributions for each of the 49 characteristic abstracted from the 1,944 drawings are tabulated in detail in Table 2. This paragraph summarizes the key observations. The median age level identified by the children for a typical older person was 71–80 years. Most children (61.2%) drew male figures. The most common physical characteristics included wrinkles (60.8%), baldness (30.6%), and hair color (24.8%). Smiles were drawn on 45.3% of figures, and 24.7% were described as happy; in contrast, 9.6% of figures were frowning, and 10.5% were described as sad, mad, or angry. Wisdom and reminiscence were often mentioned in the written descriptions (30.3 and 24.2%, respectively), whereas forgetfulness and/or dementia were infrequently described (3.2%). Canes or crutches and eyeglasses appeared in 33.5% and 28.0% of the drawings. Among health conditions, the most commonly mentioned condition was weakness (21.7%), with all other specific conditions described in < 10% of drawings. Among the recreational/vocational activities, the most commonly drawn activities were walking (12.7%), socializing (9.4%), and cooking (9.1%). Grandparents were the subject of 21.8% of drawings. For physical activity, 32.9% of drawings contained no description of any activity, and 10.3% described strenuous activity (e.g., running). Eleven characteristics were abstracted from < 5% of the drawings, and these are noted in the footnote to Table 2.

Correlations and Internal Consistency
The standardized alpha coefficient for the 49 variables was low ( = 0.37). The Spearman rho correlations between the variables were also low, with 90.2% of the 1,176 comparisons being < 0.10. The highest observed correlation was observed between the characteristics sad/mad/angry and frown (0.49). When the analyses were stratified by the gender represented in the drawing and restricted to the 34 most common characteristics, the results did not change substantially: For drawings of men and women, the standardized alpha coefficients were.30 and.35, respectively.

Exploratory Factor Analysis
Examination of a skree plot suggested a five-factor solution from a principal components factor analysis, but this explained only 21.4% of the variance in the data. The eigenvalues for the five factors were 3.09, 2.09, 2.06, 1.78, and 1.46, respectively. The loadings of the characteristics on the factors were sometimes inconsistent (i.e., one characteristic having similar loadings on more than one factor) and often weak. Factors 4 and 5 had only one characteristic loading on each factor. Separate factor analyses evaluated drawings made by boys or girls and Mexican American or European American students. These subgroup analyses did not demonstrate any differences between the factor structures for either gender or ethnic group.

We next performed exploratory factor analyses for drawings of men and women, restricted to the 34 most common characteristics. For the 1,190 drawings of elderly men, the analysis suggested a three-factor solution with eigenvalues of 2.36, 1.82, and 1.48. This solution explained 16.7% of the variance. The characteristics loading on the first factor were smile (–0.63), frown (0.57), happy (–0.51), and sad/mad/angry (0.55). Wise (0.42) and hair color (0.43) made up the second factor. Cane/crutch (–0.44) and walking (–0.51) loaded on the third factor.

For the 720 drawings of elderly women, the analysis suggested a five-factor solution with eigenvalues of 2.52, 2.01, 1.60, 1.47, and 1.41; this solution explained 26.6% of the variance. As with men, the characteristics loading on the first factor were smile (0.61), frown (–0.63), happy (0.48), and sad/mad/angry (–0.62); note that the direction of the loading is opposite of what was observed for drawings of men. Wrinkles (0.56), hair color (0.45), reminiscent (0.41), and weak (0.42) made up the second factor. Cooking (0.41) and grandparent (0.42) loaded on the third factor. Wheelchair (–0.46), activity (0.41), and walking (0.55) made up the fourth factor. The fifth factor consisted of two characteristics: pictured with others (0.54) and social activities (0.49).

The gender-specific analyses suggest three weak common factors for both men and women. The first combines characteristics of expression and personality. The second suggests associations between hair color and wisdom or reminiscence. The third common factor combines features of physical aids, activity, and walking. Among women, there were two additional factors: Cooking and grandparenting made up one, and social activities and being pictured with others made up the second. Taking all the analyses together, the observed factor structures were weak and often inconsistent. From the correlational and exploratory factor analyses, there does not appear to be any strong a priori underlying pattern of characteristics of elders as reflected by the drawings made by the children in the two schools. We therefore infer that this sample of middle school students as a group do not have clearly defined perceptions, or stereotypes, of elders.

Characteristics of Positive, Neutral, and Negative Drawings
While coding the drawings, we assigned overall ratings to each drawing as being either positive (N = 550; 28.8%), neutral (N = 939; 49.2%), or negative (N = 421; 22.0%). The odds ratios describe the likelihood that a drawing would receive a particular overall rating, given the presence of a given characteristic. Some characteristics entered all three regressions: For example, being happy was described in 24.7% of drawings; if present, the odds ratio for coding the drawing as positive was 6.84, as neutral 0.30, and as negative 0.18. In contrast, loneliness was described in only 3.2% of drawings; the presence of this characteristic yielded an odds ratio of 0.19 for a positive, 0.54 for a neutral, and 4.23 for a negative rating. In the step-forward regressions, 24, 18, and 24 characteristics were selected for the positive, neutral, and negative drawings, respectively, and accounted for 91%, 76%, and 92% of the variance in each regression (C statistic).

To reduce the number of variables further, we selected only those characteristics that increased the likelihood (odds ratio > 1.0) of a particular overall rating. These characteristics were then entered together into a logistic regression (Table 3). Thirteen characteristics were directly associated with a positive rating, 4 with a neutral, and 11 with a negative rating. This reduced set of characteristics accounted for 89%, 59%, and 85% of the variance in the positive, neutral, and negative ratings, respectively; the largest decrement in explained variance in this set of regressions was in the neutral ratings. Drawings were more likely to be coded as positive if the figure was smiling, described as happy or kind, a grandparent, pictured outdoors, and actively engaged in an activity or exercise. Negative drawings were characterized by figures being older, frowning, described as sad or grumpy, lonely or demented, weak, having trouble walking, and living with their family (loss of independence) or homeless. Neutral drawings were characterized by baldness, hair color (gray hair), walking, and watching television.

	DISCUSSION

TOP Abstract Human Figure Drawing and... Methods Results Discussion References

 
These analyses of characteristics abstracted from middle school children's drawings of elders demonstrated little evidence, a priori, of stereotypes regarding aging. Instead, the drawings demonstrated the heterogeneity that exists in children's perceptions of elders. The 49 characteristics abstracted from the drawings were not highly correlated with each other, the alpha coefficient for the internal consistency among the characteristics was low, and exploratory factor analyses failed to demonstrate a cohesive underlying structure to the variables. A strength of this study was the large sample of drawings obtained from the majority of students at two middle schools; the students themselves were not selected on the basis of any characteristics or eligibility criteria other than enrollment at the two schools. The drawing task was "open ended" in that there was no prior teaching about aging or potential stereotypes. In this way, any potential for bias that could be introduced by the investigators, the classroom teachers, or closed-ended questions with limited response categories was minimized in the children's responses. Therefore, the images should be a valid reflection of children's views at the time of making the drawing.

Although the analyses did not demonstrate a cohesive factor structure, we were able to use the abstracted characteristics to classify the drawings as positive, neutral, or negative images of elders. This classification was created by the investigators, not by the students, and reflects our own attitudes and beliefs, items that we might group as stereotypes. However, even in the subsets of negative and positive drawings, we were unable to demonstrate strong evidence for sensible clusters of characteristics that could be construed as positive, neutral, or negative stereotypes of aging. It may be that experiences that shape each child's views are so unique and varied that strong images of aging are present on the individual level but are lost when aggregated in group data. Thus, the possibility exists that some children may hold stereotypic views of elders. The catalog of characteristics documented here, their prevalence, and the positive, neutral, or negative classifications may be used in future educational research focused on improving knowledge in specific topics related to aging. For example, health promotion and disease prevention education may subsequently be reflected in children's expectations or images of their own aging.

Students who drew someone they knew, especially a grandparent, were more likely to draw a positive image than children who drew a figure from their imagination. This observation replicates the work of Couper and colleagues (1995) and underscores the importance of intergenerational contact and activities. Children who have meaningful interactions with older adults may learn skills and attitudes that will help them meet the challenges of different milestones as they progress through life. This observation underscores the importance that intergenerational activities and a life course perspective may have on the growth and development of adolescents (Jackson & Sellers, 1997).

This study has several limitations. First, although we obtained drawings from a large majority of the students in the two schools, the images produced by the students may not be a representative sample of all middle school students and the results may not generalize to other populations. Second, we did not go back to samples of students and ask them to clarify or validate what they had drawn; that is, would the creators of the drawings agree with the investigators' abstractions of characteristics and interpretations of the drawings? However, the characteristics were often readily measurable (e.g., a frown vs a smile vs another expression). Third, it is possible that our team missed abstracting additional characteristics from the drawings that might actually capture a stereotype of aging. Our approach was to abstract as much information as possible from the drawings; given the directions to the students, the likelihood that common variables or themes were missed is low. Finally, the time allotted in public school classrooms to produce the drawings may have been limited. In some classes, there may not have been enough time for the students to think about the task at hand and do their best work. Collecting drawings in public schools reflects the realities of conducting evaluative research in these settings.

As an individual moves through adolescence toward adulthood, many beliefs held as a child start to change. Analyses suggest that this life passage includes not only social experiences and regulation but changes in individual orientations toward information acquisition. Secord (1958) believes that the fluctuation in flexibility of stereotypes by children is consistent with the phase model of transitions. Flexibility appears to increase with the changes in the social environment of middle school. Then, flexibility decreases during later adolescence and entry into high school. This strengthens the justification for the use of the Positively Aging curriculum during the middle school years. Hummert (1990) assessed consistency of stereotypes across groups. This author found that stereotypes may be more consistently defined across some individuals than others and that specific traits may be more strongly associated with a particular stereotype by some individuals than others. In further research, Hummert and associates reported that older adults have more complex representations of aging than do middle-aged and young adults and that middle-aged adults have more complex representations of aging than do young adults. Thus, a very fluid pattern of stereotype development emerges, even in older ages (Hummert, Garstka, Shaner, & Strahm, 1994).

Our analyses indicate that middle school students, as a group, have not yet formed strong images regarding aging. There were no clear cohesive stereotypes of elders drawn by these children. Instead, their drawings depict elders that are diverse and multidimensional. As students transition from late childhood to adolescence, they engage in active information seeking to construct or reconstruct tentative pre-existing beliefs that the social demands of their new life phase require. Without strong stereotypes in place, there are opportunities to engage students in learning activities that offer diverse information as it relates to aging and the aging population, unfettered by prior biases (Alfieri, Ruble, & Tory, 1996). Because perceptions are limited by personal experience, it is reasonable to assume that one's perceptions of older individuals may change as experience and knowledge are gained (Secord, 1958). As stereotypes may be barriers to education (Lippmann, 1922), the findings presented here suggest that young adolescents are still at a point where instruction, including gerontological content, can be used to effectively teach about diversity in aging, successful aging, and health promotion (Pruski, Blalock, Plaetke, Murphy, Marshall, & Lichtenstein, 2003). Our hope is that adults, especially teachers, may use drawings made by middle school children to "take the pulse" of their students and use the results to explore their students' views about the connections between health and aging.

	Footnotes

 
Decision Editor: Charles F. Longino, Jr., PhD

Received for publication September 9, 2003. Accepted for publication August 12, 2004.

	References

TOP Abstract Human Figure Drawing and... Methods Results Discussion References

 

• Alfieri, T., Ruble, D. N., Tory, H. (1996). Gender stereotypes during adolescence: Developmental changes and the transition to junior high school. Developmental Psychology, 32,1129-1137.
• Altman, D. G. (1999). Practical statistics for medical research. Washington, D.C.: Chapman & Hall/CRC.
• Bassili, J. N., Reil, J. E. (1981). On the dominance of the old-age stereotype. Journal of Gerontology, 36,682-688.[Abstract/Free Full Text]
• Brewer, M., Dull, V., Lui, L. (1981). Perceptions of the elderly: Stereotypes as prototypes. Journal of Personality and Social Psychology, 41,656-670.
• Bruner, J. S., Shapiro, D., Tagiuri, R. (1958). The meaning of traits in isolation and in combination. In D. J. Mangen & W. A. Peterson (Eds.), Person perception and interpersonal behavior (pp. 277–288). Stanford, CA: Stanford University Press.
• Cortina, J. M. (1993). What is coefficient alpha? An examination of theory and applications. Journal of Applied Psychology, 78,98-104.
• Couper, D. P., Donorfio, L., Goyer, A. (1995). Final report: Images of aging: Children's attitudes. American Association of Retired Persons. Washington, DC: American Association of Retired Persons.
• Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16,99-105.
• Di Leo, J. H. (1983). Children's drawings as diagnostic aids. New York: Brunner/Mazel.
• Falchikov, N. (1990). Youthful ideas about old age: An analysis of children's drawings. International Journal of Aging and Human Development, 31,79-99.[Medline]
• Field, D., Gueldner, S. H. (2001). The oldest old: How do they differ from the old-old. Journal of Gerontological Nursing, 27,20-27.[Medline]
• Goodenough, F. L. (1926). Draw-a-Man Test. Measurement of intelligence by drawings. New York: Harcourt, Brace & World.
• Grant, M. J., Button, C. M., Hannah, T. E., Ross, A. S. (2002). Uncovering the multidimensional nature of stereotype inferences: "A within-participants study of gender, age, and physical attractiveness.". Electronic version: Current Research in Social Psychology, 8 (retrieved July 10, 2003, from http://www.uiowa.edu/grpproc/crisp/crisp.8.2htm).
• Grant, M. J., Ross, A. S., Button, C. M., Hannah, T. E., Hoskins, R. (2001). Attitudes and stereotypes about attitudes across the lifespan. Social Behavior and Personality, 29,749-762.
• Harris, D. B. (1963). Children's drawings as a measure of intellectual maturity. New York: Harcourt Brace Jovanovich.
• Harris, D. B., Pinder, G. D. (1974). The Goodenough–Harris Drawing Test as a measure of intellectual maturity of youths 12–17 years: United States. Vital and Health Statistics, series 11, no. 138.
• Hausdorff, J. M., Levy, B. R., Wei, J. Y. (1999). The power of ageism on physical function of older persons: Reversibility of age-related gait changes. Journal of American Geriatric Society, 126,17-36.
• Hickey, T., Hickey, L., Kalish, R. A. (1968). Children's perceptions of the elderly. Journal of Genetic Psychology, 112,227-235.[Medline]
• Holmbeck, G. N., Paikoff, R. L., Brooks-Gunn, J. (1995). Parenting adolescents. In Boorstein, M. H. (Ed.), Handbook of parenting, vol. 1: Children and parenting (pp. 91–118). Hillsdale, NJ: Erlbaum.
• Hosmer, D. W., Jr., Lemeshow, S. (1989). Applied logistic regression. New York: Wiley.
• Hummert, M. L. (1990). Multiple stereotypes of elderly and young adults: A comparison of structure and evaluations. Psychology and Aging, 5,182-193.[Medline]
• Hummert, M. L., Garstka, T. A., Shaner, J. L., Strahm, S. (1994). Stereotypes of the elderly held by young, middle-aged, and elderly adults. Journal of Geronotology: Psychological Sciences, 49B,P240-P249.
• Jackson, J. E. (1991). A user's guide to principal components. New York: Wiley.
• Jackson, J. S., Sellers, S. L. (1997). Psychological, social, and cultural perspectives on minority health in adolescence: A life-course framework. In Wilson, D. K., Rodrigue, J. R., & Taylor, W.C. (Eds.), Health promoting and health compromising behaviors among minority adolescents (pp. 29–49). Washington, D.C.: American Psychological Association.
• Jantz, R. K., Seefeldt, C., Galper, A., Serlock, K. (1977). Children's attitudes toward the elderly. Social Education, X,518-523.
• Kim, J. O., Mueller, C. W. (1978). Factor analysis: Statistical methods and practical issues. Quantitative applications in the social sciences, no. 07-014. London, U.K.: Sage.
• Klepsch, M., Logie, L. (1982). Children draw and tell: An introduction to the projective uses of children's human figure drawings. New York: Brunner/Mazel.
• Koppitz, E. M. (1968). Psychological evaluation of children's human figure drawings. New York: Grune & Stratton.
• Lichtenstein, M. J., Marshall, C., Pruski, L., Blalock, C., Murphy, D. L., Lee, S., et al. (1999). Positively Aging: Choices and changes. Creating interdisciplinary middle school teaching materials using examples from geriatrics and gerontology. Educational Gerontology, 25,555-569.
• Lichtenstein, M. J., Pruski, L. A., Marshall, C. E., Blalock, C. L., Lee, S., Plaetke, R. (2003). Sentence completion to assess children's views about aging. The Gerontologist, 43,339-348.
• Lichtenstein, M. J., Pruski, L. A., Marshall, C. E., Blalock, C. L., Murphy, D. L., Plaetke, R., et al. (2001). The Positively Aging teaching materials improve middle school students' images of older people. The Gerontologist, 41,322-332.
• Lippmann, W. (1922). Public opinion. Part 3: Stereotypes. New York: Harcourt, Brace.
• Malchiodi, C. (1998). Understanding children's drawings. New York: Guilford Press.
• Marks, R., Newman, S., Onowola, R. (1985). Latency-aged children's views of aging. Educational Gerontology: An International Quarterly, 11,89-99.
• Matheson, D. H., Collins, C. L., Kuehne, V. S. (2000). Older adults' multiple stereotypes of young adults. International Journal of Aging and Human Development, 51,245-257.[Medline]
• McTavish, D. G. (1971). Perceptions of old people: A review of research methodologies and findings. The Gerontologist, 11,90-101.
• McTavish, D. G. (1982). Perceptions of old people. In D. J. Mangen and W. A. Peterson (Eds.), Research instruments in social gerontology: Clinical and social psychology, vol. 1 (pp. 533–621). Minneapolis: University of Minnesota Press.
• Millstein, S. G., Petersen, A. C., Nightengale, E. O. (1993). Promoting the health of adolescents: New directions for the twenty-first century. New York: Oxford University Press.
• Mitchell, J., Wilson, K., Revicki, D., Parker, L. (1985). Children's perceptions of aging: A multidimensional approach to differences by age, sex, and race. The Gerontologist, 25,182-187.[Medline]
• Newman, S., Faux, R., Larimer, B. (1997). Children's views on aging: Their attitudes and values. The Gerontologist, 37,412-417.[Abstract]
• Palmore, E. B. (1998). The Facts on Aging Quiz. New York: Springer.
• Palmore, E. B. (1999). Ageism: Negative and positive (2nd ed.). New York: Springer.
• Parfenoff, S. H., Paikoff, R. L. (1997). Developmental and biological perspectives on minority adolescent health. In Wilson, D. K., Rodrigue, J. R., & Taylor, W. C. (Eds.), Health promoting and health compromising behaviors among minority adolescents (pp. 5–27). Washington, D.C.: American Psychological Association.
• Perdue, C. W., Gurtman, M. B. (1990). Evidence for the automaticity of ageism. Journal of Experimental Social Psychology, 26,199-216.
• Pett, M. A., Lackey, N. R., Sullivan, J. J. (2003). Making sense of factor analysis. The use of factor analysis for instrument development in health care research. Thousand Oaks, CA: Sage.
• Pinquart, M. (2002). Good news about the effects of bad old-age stereotypes. Experimental Aging Research, 28,317-336.[Medline]
• Powell, D. H. (1998). Nine myths of aging: Maximizing the quality of later life. Collingdale, PA: Diane.
• Pruski, L. A., Blalock, C. L., Plaetke, R., Murphy, D. L., Marshall, C. E., Lichtenstein, M. J. (2003). Watch your mouth! Teaching oral health and aging in the reading classroom. Educational Gerontology, 29,551-564.
• Rothbaum, F. (1983). Aging and age stereotypes. Social Cognition, 2,171-184.
• Secord, P. F. (1958). Facial features and inference processes in interpersonal perception. In D. J. Mangen & W. A. Peterson (Eds.), Person, perception, and interpersonal behavior (pp. 300–315). Stanford, CA: Stanford University Press.
• Seefeldt, C., Jantz, R. K., Galper, A., Serock, K. (1977). Using pictures to explore children's attitudes toward the elderly. The Gerontologist, 17,506-512.[Medline]
• Signori, E. I., Butt, D. S., Kozak, J. F. (1982). Attitudes toward the aged of persons under and over the age of forty. Canadian Counsellor, 16,173-179.
• Tagiuri, R., Petrullo, L., (Eds.). (1958). Person, perception, and interpersonal behavior. Stanford, CA: Stanford University Press.
• Tanaka, G. (1996). What's health got to do with it? A case for health programs in middle level schools. Midpoints: Occasional Papers, 6,1-11.
• Tuckman, J., Lorge, I. (1953). Attitudes toward old people. Journal of Social Psychology, 37,249-260.

This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation