The Formal Elements Art Therapy

Scale and “Draw a Person Picking

an Apple from a Tree”

Linda Gantt
Carmello Tabone

For more than 14 years, we have used a single-picture assessment with psychiatric patients
and have developed a rating system to measure its variables. Using these tools we conduct
both basic and applied art therapy research and clinical studies. The picture is one Lowenfeld
employed while studying children’s drawings (Lowenfeld, 1939, 1947). The instructions are
simply “Draw a person picking an apple from a tree” (PPAT). By holding the content constant
we can study the effects of a variety of demographic and psychological characteristics on the
way people draw.
Our rating system is a set of 14 scales based on global, formal attributes which we call
the Formal Elements Art Therapy Scale (FEATS) (see Table 1) and a set of Content Scales that
code for specific colors, the gender of the drawn person, and additional items beyond a per-
son, an apple, and a tree (Gantt, 2001; Gantt & Tabone, 1998).
Originally, we endeavored to measure diagnostic information using the PPAT and the
FEATS. Subsequently, we expanded the number of possible applications to a variety of special
populations as well as representative samples of both children and adults.

PICTURE COLLECTION

The art materials required are “Mr. Sketch” watercolor markers, and white drawing paper
12″ × 18″. The colors are black, brown, yellow, orange, red, purple, magenta, hot pink, tur-
quoise, blue, green, and dark green.

420
 FEATS and PPAT 421

TABLE 1. FEATS Scores on Two PPAT Drawings during ECT Treatment

FEATS scores
FEATS Scale Figure 1 Figure 2

1. Prominence of color 1.0 3.0

2. Color fit 1.5 3.0
3. Implied energy 2.0 3.0
4. Space 1.5 2.5
5. Integration 1.5 3.5
6. Logic 1.5 3.5
7. Realism 0 3.0
8. Problem. solving 0 3.5
9. Developmental level 2.0 3.5
10. Details of objects and environment 0 1.5
11. Line quality 3.0 3.0
12. Person 0 2.5
13. Rotation 2.0 5.0
14. Perseveration 4.0 5.0

Total FEATS Scorea 20.0 45.5

aOne should exercise caution in comparing the total FEATS scores from other individuals with
these above. There are several scales for which a score of 5 is associated with mania (e.g., Im-
plied Energy and Details of Objects and Environment). However, until we can establish norms
for a representative sample of nonpatients we cannot say which scores are correlated with spe-
cific diagnostic groups. At the time we do establish such norms we will be able to offer a conver-
sion formula or a range of scores so that one can interpret a total score more easily.

We hand the artist the paper so that he or she decides the orientation of the paper and say simply,
“Draw a person picking an apple from a tree.” If the person asks whether it should be a man or a
woman, we repeat the same words emphasizing the word “person.” We do not place a time limit on
doing the drawing. (Gantt & Tabone, 1998, p. 13)

We ask patients in our psychiatric hospital for a PPAT in the first art therapy session and
at discharge. Collecting the drawings at these times is crucial. We assume the pictures reflect
symptoms experienced at the time of the drawing (p. 16). For patients getting electrocon-
vulsive therapy (ECT) we obtain a baseline drawing and one after each treatment.
The ease of obtaining PPATs allows us to study a large number of different groups. We
currently have PPATs from over 5,000 psychiatric patients and approximately 1,000 adult and
child nonpatients. We are working with others to get a representative normative sample of
nonpatients for comparison.

THE ADVANTAGES OF USING THE PPAT AND THE FEATS

Lehmann and Risquez (1953) developed specific requirements for an art-based assessment:
that it be repeatable; that it could be given to anyone regardless of “artistic ability, interest,
cooperation, and intelligence”; that there be a “standardized method of rating”; and that
information could be obtained directly from the picture (p. 39). Our methods meet those crite-
ria. Yet, there are some limitations. We recognize that there can be no single art therapy
assessment, only a variety of assessments suitable for use in specific situations. Single-picture
assessments have the advantage of being easily repeated and are ideal for research. But one
422 ART-BASED ASSESSMENTS

picture gives a circumscribed view of a person’s art-making capacity and of that person’s indi-
viduality. Nonetheless, we see the advantages of using the PPAT and the FEATS as follows:

• We can easily compare one group to another.

• Our scales are based on global artistic elements.
• Our scales are correlated with specific psychiatric symptoms and diagnostic informa-
tion.
• We can apply these methods to both children and adults.
• We can measure changes over time.

Group Comparisons
If there are differences between two or more groups (whatever the group membership crite-
ria), we contend that these differences will be found in the way the group members draw
rather than what they draw. Given that the possible subject matter for art is vast, one would
have considerable difficulty testing for measurable differences in content between groups. Fur-
thermore, because much symbolism is often culture-bound, cross-cultural studies would be
problematic if the focus were only on content.
Granted, there are individual differences within any group. But, using descriptive statis-
tics (distributions and mean scores), we can arrive at a description of the group as a group.
Then, we can determine whether there are statistically significant differences between groups
that are unlikely to be by chance alone. For example, most art therapists working with hospi-
talized depressed patients would probably agree that such patients frequently make drawings
with few details and little color. Using the PPAT and the FEATS one can easily test this asser-
tion and compare that group to a group of nonpatients or to another group of patients.

Global Artistic Elements

The majority of the FEATS variables are those formal elements of great concern to artists. We
feel the research on projective drawings bogged down in a search for details that infrequently
occur and are difficult to rate. Many such details are either present or absent and are mea-
sured on nominal scales. Subtle differences are hard to capture. Global attributes such as the
use of color, integration, and implied energy can be measured on ordinal or interval scales,
making degrees of a variable easy to determine. These attributes are among the first things we
notice about a work of art. But more than that, artists are accustomed to taking in the entire
piece at a glance. As we tried to tease out measurable variables in the PPATs, we realized that
we seemed to know more than we could put into words. No doubt, many art therapists have
experienced a similar phenomenon—that they have a seemingly intuitive sense of the informa-
tion a picture conveys. We have come to recognize that this “intuition” is actually a process of
pattern matching that seemed to occur without verbal mediation (Gantt & Tabone, 1998, pp.
21–22, 52–59). Once we recognized the essential pattern we realized that it was contained in
the global formal elements.

Correlation with Symptoms and Diagnosis

The majority of the FEATS scales can be related to specific psychiatric symptoms. This is
what we call the graphic equivalent of symptoms (Gantt & Tabone, 1998, pp. 22–27). As
 FEATS and PPAT 423

we developed the FEATS we combed the literature for characteristics others associated with
particular diagnostic groups. Then, using our own clinical observations we searched the Di-
agnostic and Statistical Manual of Mental Disorders (DSM) for symptoms that could be
logically expressed in pictures. For example, if a person is severely depressed with psycho-
motor retardation, depressed mood, loss of energy, and a diminished ability to think or
concentrate (American Psychiatric Association, 1994, p. 327), an art therapist could reason-
ably expect the persons’s PPAT to have lower scores on the scales for Details of Objects and
Environment, Prominence of Color, Implied Energy, and Space than the PPATs of a
nonpatient sample.

Useful for Children as Well as Adults

We collected PPATs from 322 children in a suburban elementary school in a metropolitan
area. We applied the FEATS just as easily to these drawings as to those of adults. It is a distinct
advantage to have a rating system that applies to all ages. This makes it possible to study age-
related changes and compare our statistical results to Viktor Lowenfeld’s descriptions of de-
velopmental stages (Lowenfeld & Brittain, 1975).

Measuring Change
An advantage of the FEATS is that we can measure small changes (1 2 point). We can collect
a PPAT as part of a study on treatment response (e.g., before and after prescribing antide-
pressants, antipsychotics, or Ritalin), during ECT monitoring, or in a research design in-
volving pre- and posttesting. With patients undergoing ECT we have observed subtle
changes for the better on some FEATS variables, but these changes are not uniform. For in-
stance, a person’s drawing may be rated higher or the Person scale but lower on the Inte-
gration scale. Further studies will help us tease out the most important FEATS scales for
gauging improvement.

AN EXAMPLE OF CHANGE USING THE PPAT AND THE FEATS

Figure 1 shows an admission PPAT of a 40-year-old man, John, with a history of multiple hos-
pital admissions for schizophrenia over a 14-year period. We have followed changes in his
PPATs over this period and have more than 20 PPATs in his file. During most of his hospital
stays he was given antipsychotic medication. His usual reason for being hospitalized has been
not taking his medicine.
During one stay, John’s psychiatrist decided to give him ECT. John did a PPAT (Figure 1)
1 day before the first treatment. He selected the colors turquoise (top right-hand corner), yel-
low (small dot-like forms, purple (vertical lines), and red (two lines at right angles). John drew
most of the PPAT with the paper in a horizontal orientation and then rotated the paper verti-
cally to add the red lines. Figure 2 is John’s PPAT after one ECT session 2 days later. It has an
orange sun, a blue person, a dark green tree, and red apples. Table 1 presents the FEATS
scores on each drawing.
One year and 5 months later, John returned to the hospital. His admission PPAT was a
multicolored drawing that filled the paper, but there was no discernible person, tree, or apple
(Figure 3). After one ECT treatment John’s PPAT (Figure 4) looked similar to Figure 2. Less
424 ART-BASED ASSESSMENTS

FIGURE 1. Admission PPAT.

than 1 month later John returned to the hospital. His first PPAT resembled Figure 2. After one
ECT treatment John changed the colors of the tree to a brown trunk and green leaves and
showed the person with the apple in hand. He moved the sun to the opposite corner of the
page and made it larger.
The total FEATS scores for these last two drawings (not shown) were 39.5 and 45.5, re-
spectively. Because the hospital added other measures to its protocol for ECT monitoring we
can compare them with the FEATS scores. One day after the most recent admission John
scored 40 on the Beck Depression Inventory and 29 on the Mini-Mental Status Exam

FIGURE 2. John’s PPAT after one ECT session 2 days later.

 FEATS and PPAT 425

FIGURE 3. Admission PPAT, 1 year and 5 months later.

(MMSE). A score of 14 or higher indicates depression (Beck & Beamesderfer, 1974). The ex-
pected score for nonpatients on the MMSE is 30 (Folstein & McHugh, 1975). A day after one
ECT treatment John scored 0 on the Beck (no depression) and 28 on the MMSE (a slight de-
cline). John refused any additional ECT treatments.
Because the Beck Depression Inventory is a well-known test we could conduct a corre-
lation study to see if the FEATS scores for other patients are correlated with the Beck. If
the two measures are correlated we could claim validity for the FEATS in measuring
depression.

FIGURE 4. John’s PPAT after one ECT treatment.

426 ART-BASED ASSESSMENTS

ADDING CONTENT SCALES

Although the FEATS provides much useful information it does not answer all the questions we
have about the PPATs. For instance, two FEATS scales measure color use. The Prominence of
Color scale is concerned with the amount of color and the Color Fit scale determines whether
the artist has used colors based on the items in the task. Both are equal interval scales and
therefore are concerned with the degree of the particular attribute. Of equal interest is which
specific colors are used both for the drawing as a whole and for individual elements. For ex-
ample, as a group our control group members color the tree with a brown trunk, a green top,
and red apples. Black trunks and green or yellow apples are statistically rare. The controls also
draw the person with three or more colors. But we have many patients’ drawings with the per-
son done in only one color. Because we have seen a number of all-blue and all-yellow people
we want to know whether such color use is associated with a particular group. We added a
number of Content Scales (nominal or categorical scales) to measure such variables (Gantt &
Tabone, 1998, pp. 47–51).

RELIABILITY AND VALIDITY

A cardinal principle of any assessment is that it be both reliable and valid. Most projective
drawings fail to meet scientific muster because they cannot meet this requirement. Groth-
Marnat (1990) discusses the problems of reliability and validity in projective drawings (pp.
367–369) as well as the tendency for raters to project their own attitudes into the work (p.
370). But it is clear that he assumes, as do many researchers, that the art mirrors enduring per-
sonality traits. We do not make that assumption. Rather, because we have seen such drastic
changes in PPATs in short order, we consider these drawings a barometer of psychological
state. Few others seem to agree implicitly or explicitly with us. Treating the PPATs (or any
other art) as a reflection of the moment demands different hypotheses and imposes a specific
type of research design than has been customary.
In determining reliability we deal strictly with interrater reliability. Because we focus
on changes in state, we are not concerned with test–retest reliability. In our studies we ask
whether three raters blind to our hypotheses rate pictures in essentially the same way. Un-
less a researcher can demonstrate that his or her rating method can be applied in the same
way by others all further investigation grinds to a halt. In several studies (Gantt, 1990; Wil-
liams, Agell, Gantt, & Goodman, 1996) we demonstrated that the majority of our scales
have excellent interrater reliability, generally ranging from .88 and up. The two scales with
which we have struggled to obtain acceptable interrater reliability are Rotation and
Perseveration. This difficulty may be explained by the fact that they are not normally dis-
tributed and are rare, even in the drawings of patients. However, we are not yet ready to
discard them as both variables appear in the work of very young children and some
patients with dementia.
We are in the process of repeating the original validity study (Gantt, 1990) on a larger
sample. That research demonstrated that on 10 FEATS scales there was a statistically signifi-
cant difference between two or more groups (four diagnostic groups and a control group). Be-
cause we contend the FEATS scores can measure changes of psychological state, we will need
to do validity studies confirming the correlation of those scores with an independent measure
of clinical condition.
 FEATS and PPAT 427

REFERENCES

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.).
Washington, DC: Author.
Beck, A., & Beamesderfer, A. (1974). Assessment of depression: The depression inventory: Psychological mea-
surements in psychopathology. In P. Pichot (Ed.), Modern problems in pharmacopsychiatry (pp. 151–
169). Basel: S. Karger.
Folstein, M., & McHugh, P. (1975). “Mini-mental state”: A practical method for grading the cognitive state of
patients for the clinician. Journal of Psychiatric Research, 12, 189.
Gantt, L. (1990). A validity study of the Formal Elements Art Therapy Scale (FEATS) for diagnostic informa-
tion in patients’ drawings. Unpublished doctoral dissertation, University of Pittsburgh, Pittsburgh, PA.
Gantt, L. (2001). The Formal Elements Art Therapy Scale: A measurement system for global variables in art.
Art Therapy: Journal of the American Art Therapy Association, 18(1), 50–55.
Gantt, L., & Tabone, C. (1998). The Formal Elements Art Therapy Scale: The rating manual. Morgantown,
WV: Gargoyle Press.
Groth-Marnat, G. (1990). Handbook of psychological assessment (2nd ed.). New York: Wiley.
Lehmann, H., & Risquez, F. (1953). The use of finger paintings in the clinical evaluation of psychotic condi-
tions: A quantitative and qualitative approach. Journal of Mental Science, 99, 763–777.
Lowenfeld, V. (1939). The nature of creative activity. New York: Harcourt, Brace.
Lowenfeld, V. (1947). Creative and mental growth. New York: Macmillan.
Lowenfeld, V., & Brittain, W. (1975). Creative and mental growth (6th ed.). New York: Macmillan.
Williams, K., Agell, G., Gantt, L., & Goodman, R. (1996). Art-based diagnosis: Fact or fantasy? American
Journal of Art Therapy, 35, 9–31.