DEVELOPMENT OF TEMAS AS MULTICULTURAL TEST

About This Presentation

Title:

DEVELOPMENT OF TEMAS AS MULTICULTURAL TEST

Description:

DEVELOPMENT OF TEMAS AS MULTICULTURAL TEST – PowerPoint PPT presentation

Number of Views:139

Avg rating:3.0/5.0

Slides: 97

Provided by: hoggU

Category:

more less

Transcript and Presenter's Notes

Title: DEVELOPMENT OF TEMAS AS MULTICULTURAL TEST

1
DEVELOPMENT OF TEMAS AS MULTICULTURAL
TEST GIUSEPPE COSTANTINO, PH.D. LUTHERAN MEDICAL
CENTER LUTHERAN FAMILY HEALTH CENTERS BROOKLYN,
N.Y. FOURTHEENTH ROBERT LEE SUTHERLAND
SEMINAR HOGG FUNDATION FOR MENTAL HEALTH NOVEMBER
30-DECEMBER 1, 2006 HOUSTON, TEXAS POWERPOINT
PRESENTATION PREPARED BY ERMINIA COSTANTINO,
B.S. AMERICAN MULTICULTURAL INSTITUTE Direct all
correspondence to Dr. Giuseppe Costantino,
Director of Training and New Programs, Lutheran
Medical Center, Lutheran Family Health Centers
5121 2nd Avenue, Brooklyn, NY 11220
Telephone718 630-8919 Fax 718 630 7974 E-Mail
gcostantino_at_lmcmc.com
2

BACKGROUND
From the inception of cross-cultural
investigations using projective/narrative tests
in the 1940s,
it was observed that TAT (Murray, 1943) stimuli
had questionable relevance to individuals of
different cultures, hence, culturally sensitive
TAT pictures were developed to study such groups
as Mexicans Indians, Ojibwa Indians, Southwest
Africans, and South Pacific Micronesians (Henry,
1955). The work of Monopoli (1984, cited in Dana,
1986) indicated that culture-specific stimuli
were necessary for personality assessment of
unacculturated Hopi and Zuni Indians. However
such early efforts to provide a culture-specific
TAT pictures have not been eagerly pursued by
psychometricians
(Dana, 1986).

A second prominent issue in psychological
assessment is the validity for ethnic/racial
minority examinees of tests that have been
conceived, standardized, and validated with a
middle-class, non-minority focus (e.g. Dana
1993). While some IQ and personality tests have
been targeted for African-Americans and blacks
(Bailey Green, 1977 Thompson, 1949 Williams
Johnson, 1981) as well Latinos (Ortiz Ball,
1972 Struthers DeAvila, 1967), unfortunately
these test have not weathered critical
psychometric scrutiny (Costantino, 1992 Oakland,
1977).

A related and more fundamental concern in
clinical and school psychology is the finding
that black and Latino children evidence less
verbal fluency on traditional projective/narrative
tests compared with their nonminority
counterparts (Costantino, Malgady Vazquez,
1981).
This poses a compelling problem antecedent to
concerns with criterion-related test validity for
ethnic/racial, inasmuch as it is widely
acknowledged that content domain scores
on projective/narrative test such as achievement
motivation of inarticulate examinees are
invalidated by lack of verbal fluency (e.g.
McClelland, 1953 Smith, 1970).

In other words, to the extent that
projective/narrative test pictures are not
culturally congruent with the cognitive schema
and present experience of the examinee, lack of
identification with the characters, theme and
setting depicted in the pictures inhibits verbal
fluency and hence precludes criterion-related
validity (Anderson Anderson, 1955 Dana, 1993
Malgady Costantino, 1996).

DEVELOPMENT OF TEMAS
Given the dearth of psychological literature on
personality assessment of culturally and
linguistically diverse youngsters (Padilla, 1979)
and in the light of the pressing needs for a
projective/narrative technique oriented toward
ethnic cultures and to create culture-specific
norms for projective tests (Dana, 1986 Exner
Weiner, 1982).
The TEMAS (Tell-Me-A-Story) narrative test was
developed with culturally relevant stimuli.

There are parallel minority and nonminority
versions of TEMAS pictures (Costantino, 1978,
1987) embodying the following features
structured stimuli and diminished ambiguity to
pull for specific personality functions
chromatically attractive, ethnically relevant and
contemporary stimuli to elicit diagnostically
meaningful stories representation of both
negative and positive
intra personal functions in the form of
conflicts or dilemmas that require a solution
and objective scoring of both thematic structure
and content. The rationale for these and other
departures from traditional projective techniques
(e.g. reduced ambiguity, color) is based on
empirical research.

The presentation of culturally relevant and
familiar projective test stimuli was explored in
Thompsons (1949) Black TAT, based on the
assumption that similarity between stimulus and
examinee facilitates identification with the
characters in the pictures and therefore promotes
greater verbal fluency and self-disclosure.
Projective stimuli traditionally have been
ambiguous in order to bypass ego defenses and
allow latent psychological conflict to be more
freely expressed however, TEMAS was conceived
following more recent thinking that diminished
ambiguity and increased structure facilitates
verbal fluency and enables a more focused
understanding of the examinees personality
functioning a more reliable and valid clinical
interpretation of thematic content (Costantino,
1978, 1987 Epstein, 1966 Sobel, 1981).

Both clinicians and researchers alike acknowledge
that color has an impact on the perception of
projective stimuli (Murstein, 1963) and that
integration of color and for is considered a sign
of emotional maturity and cognitive organization
(Siipola, 1950). During the 1950s several studies
documented that chromatic TAT stimuli enhanced
verbal fluency and more accurately discriminated
between clinical and control subjects (e.g.,
Brackill 1951). Murstein (1963) has suggested
that color facilitates such differentiation of
psychiatric and normal examinees because
achromicity reinforces sadness as an affective
response to TAT stimuli. Thus, based upon this
evidence and reasoning the TEMAS stimuli were
developed in color.

The representation of psychological conflict or
problem solving in TEMAS stimuli was based on the
methodology of Kohlberg (1976), who developed
stories portraying moral dilemmas to assess the
moral development of children. Similarly, the
examinee must resolve the antithetical situations
portrayed, and the examining clinician evaluates
the adaptiveness of the resolution of conflict.
The conflicts depicted in TEMAS stimuli were
designed to evoke disclosure of specific
personality functions that are prominent in
personality theory and research, and also are key
diagnostic indicators of psychopathology.

TEMAS has the same meaning in several
languages, in English, TEMAS is an acronym for
tell me a story in Spanish, it means themes, in
Italian, TEMA, means theme. This felicitous
cross-linguistic combination of names represents
the most appropriate title for a narrative test
(Costantino, 1987 Costantino, Malgady Rogler,
1988 Ritzler, 1993). The TEMAS represents a
number of departure and improvements relative to
previous projective/narrative.

First, the test was specifically developed for
use with children and adolescents (Ritzler,
1993).
Second, the test comprises two parallel sets of
stimulus cards, one set for minorities and the
other for non-minorities, thus making it
multicultural in nature.
Third, the test abandons the construct of
pictorial ambiguity common to the TAT and
Rorschach to pull for specific conflicts and
utilizes structured stimuli with pictorial
problem solving (Costantino, Flanagan Malgady,
2001 Flanagan et. Al., 2004 Flanagan Di
Giuseppe, 1999).

Fourth, the test has normative data for black,
Puerto Rican, other Hispanic, and white children
thus increasing multicultural validity and
diminishing test bias against minorities (Dana,
1993 1998 Ritzler, 1996).
Fifth, the stimulus cards are in color, which
attracts and maintains childrens interest
(Costantino, 1978 Costantino, Malgady Rogler,
1988) and facilitates narratives of emotional
states (Costantino, Flanagan Malgady, 1995
Lubin, 1955 Murstein, 1963 Exner, 1998
Thompson Bachrach, 1951).

14
TEMAS Pictures

The TEMAS stimuli were created by a young,
upcoming artist, Phil Jacobs, who worked closely
with the test author (Costantino, 1978 1987) in
an attempt to depict a variety of psychosocial
situations. The stimuli embody a wide variety of
problematic life situations and experiences in
inner city impoverished environments, such as
familial scenes within the home, solitary
dreamlike and fantasy states, street scenes
involving peers and adults, sports activities,
and situations occurring in school settings

The antithetical nature of the situations
portrayed in the pictures enables positive or
negative feelings to be projected in stories and
manifested as adaptive or maladaptive resolutions
of the underlying genotype motivation. These
situations pull themes expressive of varying
degrees of psychopathology, ranging from severe
pathology (e.g., morbidity, suicidal ideation,
depression, impulsivity, isolation, delusion) to
highly adaptive functioning.
16
Personality Functions and Pictures Inter-Judge
Reliability.

The nine personality functions were designed and
psychometrically tested to pull for a minimum of
two to a maximum of four personality functions
however pictures Nos. 10B(Boy) G (Girl) and 20
pull for only one personality function. Specific
picture functions for each card are denoted by a
triangle in the scoring boxes of the Personality
Functions as shown in the Record Booklet. The
pull(s) for each card were empirically assessed.
In this study, Costantino, Malgady and Rogler,
1988 assessed the degree of inter-judge agreement
about card pulls in a sample of 14 clinical and
school, culturally diverse psychologists.
The percentage of agreement among the 14
clinicians ranged from 100 on card 1BG for
Interpersonal Relations to 71 on card 15 for
Achievement Motivation.

The TEMAS was designed for use with minority and
non-minority children and adolescents aged 5 to
18 it has two parallel versions minority for
Blacks and Hispanics/Latinos, and non-minority
for whites a third Asian-American version is in
the process of being validated (Yang and
Costantino, 2002). Each version is comprised of a
short form of nine cards, which can be completed
within 40-50 minutes and the long form of 23
cards, which can be completed within 100-120
minutes. The short form is usually administered
in clinical practice, which can be augmented by
two or three additional cards pulling for the
specific symptoms presented by a given child
(e.g. aggression, anxiety, sexual identification
conflicts).

Theoretical Framework
The theoretical formulation of TEMAS is largely
multi model social-cognitive and narrative
theoretical, which posits that personality
development occurs within a socio-cultural system
where individuals internalize the cultural values
and beliefs of family and society (Bandura, 1977,
1986, 1989, 1991 McAdams, 1994, 2006 Piaget
Inhelder, 1971 Sullivan, 1953). Personality
functions are learned initially through modeling
(Bandura, 1977) and are then developed through
verbal/narrative and imaginal processes (McAdams
Pals, 2006 Paivio, 1971 Piaget Inhelder,
1971
Singer Pope, 1978).

When a narrative test pictorial stimuli are
similar to the situations in which the
personality functions were originally learned,
these functions are attributed to the characters
and situations in the cards and narrated as
personal life events and life stories (Auld,
1952, 1954 Bandura, 1986 Bruner, 1986 Mancuso
Sarbin, 1983 Sarbin, 1986 Teglasi, 2001).
Narrative psychology is gaining popularity as
narrative theory (Kirkman, 2002) in the form of
both narrative assessment (e.g. Costantino, 1978
Costantino, Flanagan Malgady, 2001 Costantino,
Malgady Rogler, 1988 Teglasi, 2001) and
narrative treatment (e.g., Bracero, in press
Costantino, Malgady Rogler, 1994 Freedman
Combs, 1996 Howard, 1991

Within the context of this presentation, we will
exhibit sufficient theoretical evidence to
discuss the value of the TEMAS test as a
narrative assessment tool.
Bruner (1986) posited that there are two modes of
thought in obtaining knowledge, the paradigmatic
mode which deals with general causes and effects
in the discovery of empirical truth and the
narrative mode which deals with the way
individuals understand the vicissitude of their
own lives.

Traditionally, psychology has been using the
mechanistic laboratory and experimental methods
in obtaining knowledge about individuals. More
recently, however, Sarbin (1986) and Mancuso and
Sarbin (1983) proposed the narrative mode as a
root metaphor in psychology, which allows
psychologists to in understanding individuals
through their narratives or life histories.
Narrative theory suggests that that there is an
interrelationship between cultural narrative and
personal narrative because culture influences the
individual and in turn the individual narrative
reflects the culture in which they are told
(Kirkman, 2002)

In addition, there is a mutual relationship
between individual and cultural narrative and the
socio cognitive psychology of human development
such as Banduras (1986) and Piaget and Inhelder
(1971) Singer and Pope (1978) because those
social learning theorists must explain the
development of the self through different stages
of life span through the narrative identity of
the individual. Narrative identity is developed
within the matrix of social and cultural milieu
(Bruner, 1990) and through telling about our self
that we develop a sense of the self (Mancuso
Sarbin, 1983).

In the tradition of psychologists who have
proposed a narrative conceptualization of human
thinking and personality within a multicultural
context, McAdams and Pals (2006) write that the
complex interplay between culture and personality
functioning seems to be most apparent in
narrative identity as revealed in individuals
life stories. Life stories tell much about the
individual and, at the same time, about the
culture in which the individual lives. In
addition, Howard (1991) suggested that
storytelling is fundamental to the development of
self-identity, which he calls life story
construction. Psychopathology occurs when life
stories go awry.

The effectiveness of a culturally competent
multicultural narrative technique in personality
assessment such as the TEMAS provides a perfect
fit with the reformulation of personality as
narrative identity by McAdams and Pals (2006) and
with Howards (1991) view of psychotherapy as a
technique in story repair (Costantino, Malgady
Rogler, 1994).
Hence, TEMAS narrative test, may correspond to a
valid multicultural assessment in understanding
why life stories go wrong among culturally
diverse individuals thus linking culturally
appropriate assessment to treatment.

25
TEMAS Administration

The TEMAS administration is a reformulation of
both the traditional TAT administration (Murray,
1943) and the Rorschach Comprehensive System
administration (Exner, 1993). In order to reduce
the examiners bias, the test should be
administered in the childs dominant language by
an examiner fully familiar with the child
cultural background, and proficient in the
standardized administration of the test (Dana,
1998). Individual administration is important
because the normative data were obtained by this
method. Unlike the TAT and Rorschach, the TEMAS
cards are administered in a random order, except
for Card 1, which is administered first because
of its non-threatening content.

The administration is comprised of establishing
rapport with the examinee, General Instructions
I have several interesting pictures that I am
going to show you. Look at the persons and
places in the pictures and tell me a complete
story about each picture, one that has a
beginning and end. Specific Instructions of
Temporal Sequencing The story should answer
three questions What is happening in the picture
now? What has happened before? And what will
happen in the future? And Structured Inquiries
of all those indices missing form the stories
such as what the main character is thinking and
feeling upon resolution of the storys
conflict/problem to make a narrative scorable.

Problem Solving Scoring System
The TEMAS scoring system was developed as an
objective, problem solving (Costantino, 1987
Flanagan, 2000) and/or scoring by consensus
system (Dana, 1998). The TEMAS scoring system is
comprised of 9 personality functions, 18
cognitive functions and 7 affective functions.
Functions are scored in numbers of seconds and
minutes for Reaction Time, Spontaneous Time and
Total Time and dichotomously (1 missing, 0
not missing) when the specific function is
missing from the story vs. when it is recognized
in the story for Conflict (Problem Solution),
Sequencing, Imagination, Relationships,
Inquiries, Omissions and Transformations

The nine personality functions represented by the
TEMAS stimuli are Interpersonal Relations,
Control/Aggression, Coping/Anxiety/Depression,
Achievement Motivation, Self-Concept, Delay of
Gratification, Sexual Identity, Moral Judgment,
and Reality Testing. These are scored on a
scale from
1 (highly maladaptive) to 4 (highly adaptive)
when these functions are present in the stories
whereas those designed personality functions
which are not present in the narratives are
scored as N (the function not pulled) in a
marked triangle on the Scoring Sheet. Some
specific examples of the scoring system are
providing below.
A scoring of 1 for any Personality Function
indicates the presence of a highly maladaptive
action or resolution for a particular TEMAS card

For example, inclusions of murder, rape, assault,
physical fighting with the intent to harm in
TEMAS stories are always scored 1 for
Interpersonal Relations (IR), Aggression (AGG),
and Moral Judgment (MJ).
A score of 2 for any Personality Function
reflects a maladaptive response or resolution.
The following responses calls for a 2
Interpersonal Relations (IT) children fail to
obey their parents Moral Judgment (MJ) children
fail to admit wrong doing, without feeling
guilty, and get away with it Aggression (AGG)
(Control of) conflicts are resolved by fighting.

A score of 3 represents a moderately adaptive
resolution.
Responses scored as 3 include the following
examples a child compromises between obeying a
parental command and playing with his/her friends
following compliance (IR and DG) two siblings
accept responsibility for wrong doing instead of
fighting (IR, AGG, RT) money is saved for a time
and then is spent to buy a larger reward (DG).
A score of 4 represents a highly adaptive
response or resolution.
The child must perceive the intended conflict and
solve the problem in a mature, socially adaptive
manner. A score of 4 includes the following
types of responses a child rejects the notion of
cheating as being antithetical to learning,
thinks only of wrongdoing (if depicted in the
picture) but rejects it because is socially
unacceptable.
An N-Score (N) is entered in the appropriate
triangle on page 2 of the Record Booklet whenever
an examinee fails to verbalize an expected theme,
which is pulled by the specific card, but is not
explicitly present in the narrative. Conversely,
a specific personality function, which was
designed to be pulled by a particular card, is
not present in the corresponding story, is scored
in the appropriate box outside the triangle.

Standardization Sample
TEMAS was standardized on a sample of 642
children (281 males and 361 females) from public
schools in the New York City area. These
children ranged in age from 5 to 13 years, with a
mean age of 8.9 years (SD 1.9). The total
sample represented four ethnic/racial groups
Puerto Ricans, other Hispanics, Blacks, and
Whites. Data on the socioeconomic status (SES) of
the standardization sample indicate that these
subjects were from predominately lower and
middle-income families.

Quantitative Scales and Qualitative Indicators
The nature of the distribution of some TEMAS
functions made it impractical to convert them to
standard scores, because scores other than zero
were rare in the standardization sample. These
functions were designated Qualitative
Indicators. The TEMAS functions that had
relatively normal distributions were designated
Quantitative Scales. Qualitative indicators
are only scored on certain Cognitive and
Affective functions.

Stratification of the Standardization Sample
In the standardization sample, significant
correlations of low magnitude were found between
age and many of the TEMAS functions.
Correlations ranged from .01 to -.52 (see Table
1). Although these correlations are small, it is
believed that they reflect real developmental
trends in the childrens cognitive, affective,
and personality trends, while still retaining
respectable sample sizes, age was collapsed into
three age-range groups 5 to 7 year olds, 8 to 10
year olds, and 11 to 13 year olds

Short Form
Means and standard deviations for the short form
were derived by extracting the scores of the 9
cards from the 23-card protocols of the
standardization sample. The correlations between
the 23-card long form of TEMAS and the 9-card
short form for each function were computed
separately for the total sample and for each
ethnic/racial group (see Table 2). The
correlation between the long form and the short
form was uniformly high across samples. The
median correlation between forms was .81 for the
Total sample, .82 for Whites, .80 for Blacks, .80
for Puerto Ricans, and .81 for other Hispanics.

35
(No Transcript)
36
(No Transcript)
37
(No Transcript)
38
TABLE 2. Correlation Between TEMAS Long and Short
Forms
39
Derivation of Standard Scores

To enable users to directly compare scores within
a single protocol for the purpose of personality
profiling, or profile a clinical clients
psychological strengths and needs, raw scores of
Quantitative Scales were converted to
standardized unit normal T-Scores (Mean 50, SD
10). To minimize irregularities in the raw
score distribution, an analytic smoothing
technique was also used (Cureton Turkey, 1951).
Because it was inappropriate to transform raw
scores of the Qualitative Indicators to standard
scores, critical levels based on raw score
distributions have been developed. Based on
expert clinical evaluation, the Qualitative
Indicators should be named Clinical Indicators.

Research Evidence
Reliability
Interrater Reliability. The first interrater
study (Costantino, Malgady Rogler, 1988)
reliability conducted was conducted with 27
Hispanic 26 Black children were drawn at random
from the sample of 73 Hispanics and 42 Blacks
described in the section on internal consistency
of the Long form. The interrater agreement for
Personality Functions in the first interrater
reliability study ranged from 31-100.
Interrater reliabilities in scoring
Total Omissions and Transformations are generally
moderate-to-high for both the Hispanic and the
Black protocols (See Table 3). Little difference
is evident as a function of ethnic/racial group.

Raters generally showed greater agreement in
scoring Omissions than Transformations. Although
illogical synthesis and integration of ideas
regarding resolution of conflict, sequencing,
imagination, and relationships, generally
occurred rarely in both samples, available
estimates of correlations are suggestive of
moderate-to-high interrater agreement.
For the Affective functions, the pattern of
correlation between raters is generally high,
with no substantive differences between the
Hispanic and Black samples. With respect to the
Personality functions, correlations are
low-to-moderate for Reality Testing and sexual
identity in the Hispanic sample and substantially
higher for the remaining function

Contrary to the pattern of internal consistency
reliability estimates, the interrater
reliabilities obtained for Hispanics are
generally higher than for Black sample.
Interrater reliability was also
estimated in a recent study of the non-minority
version of the TEMAS short form (Costantino,
Malgady, Casullo Castillo, 1991). Two
experienced clinical psychologists (one with
extensive training in scoring TEMAS and the other
a newly trained scorer) independently rated 20
protocols. The results of this study indicated a
high interrater agreement in scoring protocols
for Personality Functions ranged from 75-95.
The mean level of interrater agreement was 81,
and in no cases were the two independent ratings
different by more than one-rating scale-point. It
is important to clarify that whereas, in the
recent study, the interrater agreement ranged
from 75-95. The explanation for this
discrepancy is that during the first study, which
was conducted in 1983, the TEMAS scoring system
was still undergoing changes, whereas in the
second study, which was conducted in 1987, the
scoring system and the instructions were
completely formulated.

Internal Consistency
In this context, internal consistency refers to
the degree to which individual TEMAS cards are
interrelated in measuring particular functions.
It should be noted that TEMAS cards are analogous
to test items, and it is widely known that
reliability is a function, in part, of test
length. Therefore, functions pulled by
relatively few cards (e.g., sexual identity) will
tend to show low reliability for this reason
alone.
Long Form. Internal consistency reliabilities
of the TEMAS functions were derived using a
sample of 73 Hispanic and 42 Black children (see
Table 3). The internal consistency reliability
coefficients for the Hispanic sample ranged from
.41 for Ambivalent, an affective function, to .98
for Fluency, a cognitive function, and had a
median value of .73

For the Black sample, coefficients ranged from
.31 for Setting Transformations to .97 for
Fluency, with a median of .62.
Reaction Time, Fluency, and Total Time
demonstrated high levels of internal consistency
in both Hispanic and Black samples. However, in
general, Omissions and Transformations of
perceptual details (Main Character, Secondary
Character, Event, and Setting) had lower
magnitudes of internal consistency than other
TEMAS functions in both samples. This may be
attributable to the fact that these two
functions, being clinical scales, tend to occur
less frequently, in nonclinical children
(Costantino, Colon-Malgady, Malgady, Perez,
1991). The internal consistency reliabilities
for Omissions and Transformations were uniformly
lower for Blacks than for Hispanics.

Conflict, Imagination, and Relationships
demonstrated moderate-to-high internal
consistency reliability in both ethnic/racial
groups. The alpha coefficient for Sequencing, a
cognitive function, was moderately high in the
Hispanic sample but low in the Black sample.
With respect to affective functions, reliability
estimates in the Hispanic sample were highest for
Happy, Sad, Angry, and Fearful, whereas in the
Black sample, the highest reliability was evident
for Sad, Angry, Neutral, and Ambivalent.
With respect to personality, pictures pulling for
Interpersonal Relations, Aggression, and Moral
Judgment showed the highest levels of internal
consistency in the Hispanic sample, whereas
Anxiety/Depression, Achievement Motivation, Delay
of Gratification, Self-concept, Sexual Identity,
and Reality Testing had a low-to-moderate
reliability.

For Blacks, alphas were again uniformly lower
than for Hispanics, with the highest
reliabilities associated with Aggression and
Moral Judgment. Low reliabilities for the
personality functions may be due partially to the
fact that personality function scores are based
on relatively few TEMAS cards.
The coefficient alphas for the standardization
sample, differentiated by ethnic/racial group
membership for the long form, were, for the most
part, in the moderate range, with a median alpha
of .83 for the Quantitative Scales for the Total
Sample. On these functions, the median
reliability ranged from .80 for Black children to
.69 for other Hispanic children.

On the Short form, alphas were generally lower,
with a median reliability of .68 for the Total
sample on the Quantitative scales. Reliability
coefficients for ethnic/racial groups on these
functions ranged from a median coefficient of .65
for the White sample to .54 for the Black sample.
Reliability coefficients on the Qualitative
Indicators were lower, due, in large part, to the
nonmetric nature of the scoring system used with
these scales.

Test-Retest Reliability (Short Form)
Test-retest stability of the TEMAS functions was
computed for the Short Form by correlating the
results of two administrations, separated by an
18-week interval. The sample used in this study
consisted of 51 subjects chosen at random from
the 210 Puerto Rican students screened for
behavior problems. Results indicated that TEMAS
functions exhibited low-to-moderate stability
over an 18-week period (see Table 3). The eight
TEMAS functions with significant test-retest
correlations were fluency, event transformations,
conflict, relationships, happy, ambivalent,
anxiety/depression, and sexual identity.

First, test-retest correlations may be
lower-bounds estimates of reliability in this
case because different raters were employed at
pre and post-testing. Therefore, they include
error variance due to interrater reliability.
Second, the indicators of this instrument have
limited range and hence, the correlation may be
attenuated. Third, the test interval was over
four months, a lengthy period for reliability
estimation, especially the younger the children,
who are undergoing maturational changes (i.e.,
the trait itself may be changing developmentally
over time).

50
(No Transcript)
51
(No Transcript)
52

Validity
Face Validity
TEMAS pictures were designed to pull for
specific personality functions based upon the
nature of the psychological conflict represented
in each picture. As previously described in the
Scoring section, all TEMAS pictures are scored
for at least two and not more than four
personality functions. Face validity was
established by a panel of bilingual/bicultural
expert clinicians. The percentage of agreement
among 14 clinicians on the likely pulls of each
TEMAS care revealed very high agreement
(71-100) across the pictures, thus confirming
the pulls scored for specific personality
functions.

Criterion-Related Validity
A group of 210 Puerto Rican children screened for
behavior problems were administered a number of
measures along with the TEMAS, and their adaptive
behavior in experimental role-playing situations
was observed and rated by psychological
examiners. The measures administered included
the Sentence Completion Test of Ego Development
(SCT Loevinger Wessler, 1970) or its Spanish
version (Brenes-Jette, 1987) the Trait Anxiety
Scale of the State-Trait Anxiety Inventory for
Children (STAIC Speilberger, Edwards, Lushene,
Montouri, Platzek, 1973) or its Spanish
version, Inventario de Ansiedad Rasgo-Estado Para
Ninos (Villamil, 1973) the Teacher Behavior
Rating Scale (TBR Costantino, 1980), and the
parallel Mother Behavior Rating Scale (MBR
Costantino, 1980) in both English and Spanish.
Finally, the children participated in four
experimental role-playing situations, designed to
elicit adaptive behavior.

Results of multiple regression analyses indicated
that TEMAS profiles significantly predicted ego
development (SCT), R. 49, plt. 05 disruptive
behavior (DIS), R. 51, plt. 05 and aggressive
behavior (AGG), R. 32, plt. 05. However, the
multiple correlations for predicting trait
anxiety were not significant. TEMAS functions
accounted for between 10 (for AGG) and 26 (for
DIS) of the variability in scores on the
criterion measures. Predictive validity was
established using hierarchical multiple
regression analysis to assess the utility of
TEMAS profiles for predicting post-therapy scores
(N123) on the criterion measures, independent of
pretherapy scores. In the first step of the
hierarchy, the pretherapy score on a given
criterion measure was entered into the regression
equation, followed in the second step by a
complete TEMAS psychotherapy profile.

Results of these analyses showed that pretherapy
TEMAS profiles significantly predicted (p lt .05)
all therapeutic outcomes, ranging from 6 to 22
variance increments, except for observation of
Self-concept of Competence. Variance explained
in the outcome measures were the Sentence
Completion Test of Ego Development (14) Trait
Anxiety Inventory for Children (22) Conners
Behavior Rating Scale (6) and observational
tasks measuring delay gratification (20)
disruptive behavior (17) and aggression (14).

Population Generalizability
Clinical information provided by TEMAS
profiling of personality, cognitive and affective
functioning is generalizable with a high degree
of confidence to multiple ethnic/racial groups in
the 5 to 13 year-old range. Gender differences
are relatively inconsequential based upon
statistical comparisons of differential
reliability and validity, as well as normative
performance. Several multicultural and
cross-cultural studies to be a valid assessment
measure with culturally divers youngster in the
USA and other foreign countries (Cardalda et al.
1998 Costantino, Malgady, Casullo, Castillo,
1991 Costantino Malgady 1996, 1999 Dana,
1996 Carnabuci, 2000 Fantini, 2005 Sardi,
2000 Summo, 2000 Sulfaro, 2000 Walton,
Nuttall, Vazquez-Nuttall, 1997 Yang, Kou
Costantino, 2002)

Language of TEMAS administration has been
thoroughly researched in English, Spanish,
Italian and Mandarin. Although the test has been
used on a pilot basis in other languages,
including Italian, Portuguese, Cantonese, and
Mandarin, insufficient data are available to
confirm the psychometric properties of the
instrument with these and other non-English
languages. Clinical status of examinees is
another important consideration in the
generalizability of TEMAS scores. For the most
part, participants in TEMAS research have fallen
into three categories normal, borderline
behavior problems, and ADHD clinical cases. No
research has been conducted with children
experiencing other specific DSM disorders of
childhood and adolescence. More research in this
area is warranted.

Overall Conclusions
The rationale for developing TEMAS by departing
from the traditional features of the
projective/narrative techniques by such
modifications as the use of color, reduced
ambiguity and an objective scoring is based on
empirical research (Dana, 1993, 2000 Ritzler,
1993, 1997). Multicultural and cross-cultural
research indicates that the test shows validity
and clinical utility with culturally diverse
children and adolescents in the USA, Puerto Rico,
Argentina, Peru, Italy and Taiwan

The TEMAS scoring system presents a systematic,
and a psychometrically reliable and valid system
to score the complexity of personal narratives.
Its clinical utility is based on the design of
structured pictures which depict characters that
promote identification between the storyteller
and the stimuli and on the development of a
scoring system which closely captures the most
salient aspects of the personal narratives thus
showing both emic and etic validity (Cardalda et
al. 1998, 2005 Costantino, Malgady, Casullo,
Castillo, 1991 Costantino Malgady 1996, 1999,
200b Costantino, Malgady Dana, in Press Dana,
1993 Carnabuci, 2000 Dupertuis et al., 2005
Fantini, 2005 Sardi, 2000 Summo, 2000 Sulfaro,
2000 Walton, Nuttall, Vazquez-Nuttall, 1997
Yang, Kou Costantino, 2002).

Fowers and Davidov (2006) reformulated
Multiculturalism in psychology by emphasizing
that scientific, social, ethical and moral
principles are necessary and sufficient tenets
for embracing, learning and practicing
multiculturalism, with the ethical and moral
principles providing the strongest link.