The Adult Sensory Objective.

This article describes a series of studies designed

to evaluate the reliability and validity of the Adult Sensory
Profile.
Profile: Measuring Method. Expert judges evaluated the construct validity
of the items. Coefficient alpha, factor analysis, and correla-
Patterns of Sensory tions of items with subscales determined item reliability,
using data from 615 adult sensory profiles. A subsample of
20 adults furnished skin conductance data. A heterogeneous
Processing group of 93 adults completed the revised Adult Sensory
Profile, and item reliability was reexamined.
Results. Expert judgment indicated that items could be
categorized according to Dunn’s Model of Sensory
Processing. Results suggested reasonable item reliability for
Catana Brown, Nona Tollefson, Winnie all subscales except for the Sensation Avoiding subscale.
Dunn, Rue Cromwell, Diane Filion Skin conductance measures detected distinct patterns of
physiological responses consistent with the four-quadrant
model. Revision of the Adult Sensory Profile resulted in
Key Words: behavior • neurological model • improved reliability of the Sensation Avoiding subscale.
sensation Conclusion. The series of studies provides evidence to
support the four subscales of the Adult Sensory Profile as
distinct constructs of sensory processing preferences.

Brown, C., Tollefson, N., Dunn, W., Cromwell, R., & Filion, D.
(2001). The Adult Sensory Profile: Measuring patterns of sensory pro-
cessing. American Journal of Occupational Therapy, 55, 75–82.

S
ensory processing is recognized as a domain of con-
cern in occupational therapy and is included as a per-
formance component in the third edition of the
Uniform Terminology for Occupational Therapy (American
Occupational Therapy Association, 1994). Occupational
therapists typically assess sensory processing with measures
of behavioral response to sensation. Although there are a
number of measures available to assess sensory processing,
the majority were developed for children. For example, in
a recent review of sensory processing tests used by occupa-
Catana Brown, PhD, OTR, is Associate Professor, Department of tional therapists, all measures listed were for children
Occupational Therapy Education, University of Kansas Medical (Kohlmeyer, 1998).
Center, 3033 Robinson, 3901 Rainbow Boulevard, Kansas City,
The lack of adult measures is unfortunate, given evi-
Kansas 66160; [email protected].
dence that sensory processing concerns in children also are
Nona Tollefson, PhD, is Professor, Department of Psychology experienced by adults. When selected subtests of the
and Research in Education, University of Kansas. Southern California Sensory Integration Test (SCSIT) were
administered to adults with psychiatric disabilities (Falk-
Winnie Dunn, PhD, OTR, is Professor, Department of Kessler, Quittman, & Moore, 1988), significant relation-
Occupational Therapy Education, University of Kansas Medical ships were found between the SCSIT and the neurological
Center. dysfunction present in the sample. A study of sensory
Rue Cromwell, PhD, is Distinguished Professor, Department of defensiveness in persons with developmental disabilities
Psychology, University of Kansas. showed that four of six behaviors were equally prevalent in
children and adults (Baranek, Foster, & Berkson, 1997).
Diane Filion, PhD, is Assistant Professor, Department of Another study proposed a conceptual framework for sen-
Psychology, University of Missouri in Kansas City. sory defensiveness in adults after identifying common
themes related to response and coping to sensory stimuli
This article was accepted for publication January 10, 2000.
(Kinnealey, Oliver, & Wilbarger, 1995).
The American Journal of Occupational Therapy 75
Downloaded from http://ajot.aota.org on 09/24/2019 Terms of use: http://AOTA.org/terms
 Clearly, adult measures of sensory processing are war- four-quadrant model. The ninth factor, fine motor/percep-
ranted and needed. A measure that is based on a theoreti- tual, is not included in the model because it includes
cal model and includes all sensory modalities also is behaviors across all continua.
desired. This article describes the process of developing the Further explanation of Dunn’s Model of Sensory
Adult Sensory Profile, which is based on Dunn’s (1997) Processing follows, with a clarification of the two continua
Model of Sensory Processing and the procedures used to and the resulting quadrants. The two poles of the neuro-
examine the psychometric properties of the measure. logical threshold continuum are low and high. A low neu-
rological threshold indicates that the person requires low-
Background intensity stimuli for neurons to fire and for the person to
Sensory History as a Behavioral Measure of Sensory react. Conversely, a person with a high neurological thresh-
Processing old requires high-intensity stimuli or takes longer to react
One way to measure sensory processing from a behavioral to the same stimuli. The two poles of the behavioral
perspective is to obtain a sensory history. Advantages of response continuum are accordance and counteract.
sensory histories over other measures include ease of admin- Accordance indicates that the behavior corresponds with
istration and contextual relevance (Dunn, 1994). Sensory the neurological threshold. Responding in accordance with
histories are contextually relevant because they allow evalu- a low threshold means that the person easily and quickly
ation of behaviors in the natural environment rather than recognizes and responds to sensory stimuli. Conversely,
the commonly used alternative of observation of perfor- responding in accordance with a high threshold means that
mance in a clinical setting. Additionally, sensory histories the person takes longer to respond to or misses available
allow the person or family member who is the focus of the sensory stimuli. Counteracting behaviors indicate that the
history taking to be an active participant in the evaluation. person is responding contrary to the neurological thresh-
old. When a person counteracts a low threshold, the behav-
The Sensory Profile iors involve avoiding sensory stimuli because he or she per-
ceives even low-intensity stimuli and easily is inundated or
The Sensory Profile is a sensory history developed as a mea-
sure of children’s responses to everyday sensory experiences overwhelmed by the input. On the other hand, counter-
(Dunn, 1994). The original version of the measure is divid- acting a high threshold entails pursuit or immersion in sen-
ed into six sensory categories (auditory, visual, taste/smell, sory stimuli because the person is attempting to meet a
movement, body position, touch) and two behavioral cate- threshold that requires intense stimuli.
gories (emotional/social, activity level) for a total of 125 Dunn (1997) described each quadrant resulting from
items. A principal components factor analysis of the the interaction of the neurological threshold continuum
Sensory Profile was conducted on the basis of the respons- and the behavioral response continuum. The first quad-
es of 1,115 children 3 to 10 years of age who were typical- rant, sensitivity to stimuli, represents behaviors in accor-
ly developing (Dunn & Brown, 1997). The resulting factor dance with a low neurological threshold. Distractibility,
structure did not suggest categories of sensory modalities difficulty screening stimuli, and discomfort with sensation
but rather patterns of behavioral responsiveness. This factor characterize this quadrant. The second quadrant, counter-
analysis led to the development of Dunn’s (1997) Model of acting a low neurological threshold, is labeled sensation
Sensory Processing, which characterizes four different sen- avoiding and includes behaviors that limit exposure to stim-
sory processing tendencies. uli. The third quadrant, low registration, reflects responses
in accordance with a high neurological threshold. This
Dunn’s Model of Sensory Processing
quadrant includes a disregard of or slow response to sensa-
Dunn’s Model of Sensory Processing explains behavioral tion. The final quadrant, sensation seeking, is a counterac-
responses to sensation in children. The findings from the tive response to high neurological threshold and encom-
factor analysis of the Sensory Profile (Dunn & Brown, passes pleasure derived from rich sensory environments and
1997) contributed to the model’s conceptualization on the behaviors that create sensation.
basis of the relationship between a neurological threshold Dunn’s Model of Sensory Processing depicts sensory
continuum and a behavioral response continuum. The fac- processing preferences as stable traits; therefore, applying
tor analysis resulted in nine factors: (a) sensory seeking, (b) the model to adults is inherently reasonable. The extension
emotionally reactive, (c) low endurance/tone, (d) oral sen- of the model led to the development of the Adult Sensory
sory/sensitivity, (e) inattention/distractibility, (f) poor reg- Profile as a measure of adult responses to everyday sensory
istration, (g) sensory sensitivity (h) sedentary, and (i) fine experiences. The purpose of the current article is to report
motor/perceptual. From these data, Dunn developed a the results of a series of studies designed to evaluate the reli-
four-quadrant model based on the intersections of a neu- ability and validity of the Adult Sensory Profile as a mea-
rological threshold continuum and a behavioral continu- sure of the four patterns of sensory processing described in
um. Figure 1 illustrates the grouping of factors into the Dunn’s model.

76 January/February 2001, Volume 55, Number 1

Downloaded from http://ajot.aota.org on 09/24/2019 Terms of use: http://AOTA.org/terms
 BEHAVIORAL RESPONSE CONTINUUM Procedure. The researcher defined the four-quadrant
Accordance Counteract model of sensory processing for the judges, encouraging
clarifying questions. Once familiar with the concepts, each
LOW REGISTRATION SENSATION SEEKING
judge received the 60 items written individually on a set of
Factor 3 (low endurance/tone) Factor 1 (sensation seeking) shuffled index cards. The judges were instructed to sort the
NEUROLOGICAL THRESHOLD CONTINUUM
High

Factor 6 (poor registration)

Factor 8 (sedentary)
cards into the four quadrants of sensory processing. They
did not know how many items were developed for each
Expected physiological response to Expected physiological response category. The result from each judge was four sets of cards
sensation is a weak response (due to sensation is a weak response
to high threshold) and quick (due to high threshold) and representing sensory sensitivity, sensation avoiding, low
habituation (due to accordance slow habituation (due to coun- registration, and sensation seeking. After the card sorting,
behavior that continues to limit teract behavior that pursues sen-
response) sation) the researcher read each item aloud and asked each judge to
identify the quadrant to which the item was assigned.
Analysis. The researcher recorded each judge’s respons-
SENSORY SENSITIVITY SENSATION AVOIDING es to determine the proportion of agreement. Item accept-
ability was set at 75% agreement among the judges
Factor 4 (oral sensory/sensitivity) Factor 2 (emotionally reactive)
Factor 5 (inattention/ Factor 8 (sedentary when moti- (Crocker & Algina, 1986).
distractibility) vation is to keep away from Results. On the basis of the review of 60 items by eight
Factor 7 (sensory sensitivity) sensory experiences)
Low

expert judges, all but one item was sorted accurately. Thus,
Expected physiological response to Expected physiological response to the items could be categorized according to the extant the-
sensation is strong response (due to sensation is a strong response (due ory. Accurate sorting in this case meant that at least seven
low threshold) with slow habitu- to low threshold) with quick
ation (due to accordance behav- habituation (due to counteract of the eight judges sorted the item as it was intended. This
ior that involves a sustained behavior that withdraws from degree of agreement exceeded the previously set criterion of
recognition of available sensa- sensation)
tion) 75%. The problematic item (miscategorized by three
judges), “I need to hold onto walls or banisters to go down
Figure 1. Sensory Profile (child version) factors and relation- steps” (a low registration item), was revised to read, “I am
ship to Dunn’s (1997) Model of Sensory Processing.
unsure of footing when walking on stairs.” Judges were able
Method and Results: Steps 1 Through 4 to categorize the item correctly once it was reworded.
Design
Step 2: Item Reliability and Factor Analysis
The study was conducted in a series of four steps; the
Purpose. In this step, the researchers examined the consis-
method and results for each step will be presented togeth-
tency of the psychometric properties of the Adult Sensory
er. The first step of instrument development consisted of a
Profile and the quadrants of Dunn’s Model of Sensory
review of the items on the Sensory Profile. Non-age-specif-
ic items were maintained and new items were created to Processing. Specifically, the questions addressed were as fol-
ensure an adequate number of items representing the four lows: Are the quadrant subscales internally consistent? Do
quadrants of Dunn’s Model of Sensory Processing and the the items correlate or load on the intended subscale? Do
sensory modalities. A 60-item measure, with 15 items for the items result in factors compatible with the four-quad-
each quadrant, was created. Subsequent instrument devel- rant model?
opment included a review of the items by an expert panel, Participants. A total of 615 adults completed the Adult
item reliability and factor analysis of typical adult respons- Sensory Profile. Psychology students and occupational ther-
es, and an analysis of construct validity using Adult Sensory apy students from a large midwestern university (n = 476)
Profile scores and skin conductance measures. All steps of were recruited through their departments. The researchers
the study were approved by the university’s Institutional obtained additional participants through the Sensory Profile
Review Board. mailing list, which consists primarily of practicing occupa-
tional therapists who have shown an interest in the measure.
Step 1: Expert Panel The researchers sent letters requesting that each list member
Purpose. The purpose of the first step was to examine the complete the Adult Sensory Profile and that they ask anoth-
face validity of the Adult Sensory Profile. This step was er adult of the opposite gender to complete the form. This
accomplished by determining whether persons familiar effort provided a larger sample, a more even gender ratio,
with Dunn’s Model of Sensory Processing could accurately and a wider age range. The total sample included 38.8%
categorize items according to the intended quadrant. men and 61.2% women, with a mean age of 30.7 years (SD
Participants. Eight expert judges comprised the panel, = 13.3, range = 17–79 years).
which consisted of three faculty members from an occupa- Procedure. All participants were informed via a cover
tional therapy department and three graduate students and letter that completion of the Adult Sensory Profile indicat-
two psychology faculty members from a research seminar. ed consent to participate in the study. Participants inde-

The American Journal of Occupational Therapy 77

Downloaded from http://ajot.aota.org on 09/24/2019 Terms of use: http://AOTA.org/terms
 pendently completed the Adult Sensory Profile and creates some doubt about whether these two factors can be
returned the form to the researcher. interpretable as separate dimensions—one that reflects a
Analysis. Several analyses were conducted to obtain response in accordance with a low threshold (sensory sensi-
information on the reliability of individual items and sub- tivity), and one that reflects a response that counteracts a
scales (i.e., Sensory Sensitivity, Sensation Avoiding, Low low threshold (sensation avoiding). Three additional prob-
Registration, Sensation Seeking) of the Adult Sensory lems with the factor structure occurred: (a) Some items
Profile. Coefficient alpha was calculated as an internal con- loaded contrary to expectation (i.e., two sensitivity and six
sistency estimate for each 15-item subscale (Green, Salkind, avoiding items loaded on the low registration factor, and
& Akey, 1997). Next, each item on a subscale was correlat- one avoiding item loaded on the sensation seeking factor);
ed with the total score for the subscale, using Pearson prod- (b) one item loaded on both Factors 1 and 3 (i.e., “I stay
uct-moment correlations (Green et al., 1997). Finally, the away from crowds”); and (c) eight items did not load on
60 items of the Adult Sensory Profile were analyzed with a any factor. Factor 1 accounted for 10.89% of the variance;
principal component factor analysis (Gorsuch, 1983). Factor 2 accounted for 7.85% of the variance; Factor 3
Results. For item reliability, first, the Sensory Sensitivity accounted for 7.59% of the variance; and Factor 4 account-
subscale yielded a coefficient alpha value of .81, with alpha ed for 6.18% of the variance. The total variance accounted
values of .66 for the Sensation Avoiding subscale, .82 for for by the four factors was 32.5%.
the Low Registration subscale, and .79 for the Sensation When the information from the internal consistency
Seeking subscale. The Sensation Avoiding subscale lacked analysis, item correlations with subscales, and factor analy-
the internal consistency desired in attitude scales; however, sis is taken together, the Sensation Avoiding subscale does
the remaining three subscales appear reasonably internally not exhibit adequate psychometric properties. It has ques-
consistent. tionable internal consistency and lacks factorial validity
Second, each item on a subscale was correlated with (Green et al., 1997). Many of its items relate to other sub-
the total score for the subscale, using Pearson product- scales. To improve the Sensation Avoiding subscale, its
moment correlations. All items on the Sensory Sensitivity items were examined to determine characteristics that dis-
(r = .32–.56), Low Registration (r = .33–.56), and tinguished between items that were performing well and
Sensation Seeking (r = .26–.50) subscales correlated most items that were performing poorly. The discrepancy
highly with their intended subscale. In contrast, the between the two sets suggested a revision such that items
Sensation Avoiding subscale (r = .11–.52) had 11 items on the Sensation Avoiding subscale reflected more deliber-
correlating highest on a subscale other than its own. These ate avoidance behaviors. Problematic items were revised,
items had their highest correlations with the total scores on deleted, or rewritten so that the original design of 15 items
the Sensory Sensitivity or Low Registration subscales. per subscale was maintained. Conditions that led to
Next, the 60 items of the Adult Sensory Profile were changes included the items that correlated highest with an
analyzed with a principal component factor analysis. The cri- unintended subscale, items with no loadings greater than
teria used to determine the number of factors to rotate .40 in the factor analysis, cross-loading of items, and clari-
included the scree test, the a priori hypothesis of four factors, fying comments written on the measure by participants.
and the interpretability of the factor solution. The criterion
of eigenvalues greater than 1 was not useful because it would Step 3: Construct Validity—Physiological Response and the
Adult Sensory Profile
have yielded an unwieldy 16 factors, with most factors con-
tributing less than 4% of variance. The scree test suggested a Purpose. The purpose of this step was to examine construct
four-factor solution, and the four-factor solution led to the validity of the Adult Sensory Profile by determining
most interpretable results. Consequently, four factors were whether a strong preference for a sensory processing pattern
rotated, using a Varimax rotation procedure. resulted in different physiological response patterns as mea-
The four-factor solution is generally supportive of sured by skin conductance amplitude of response and trials
Dunn’s Model of Sensory Processing, with some stipula- to habituation. The a priori hypotheses were as follows:
tions. Item loading of .40 or more was set as the criterion 1. Persons with low thresholds (sensory sensitivity
of magnitude to be a meaningful item for a particular pat- and sensation avoiding) would have a greater
tern of sensory processing. Of the four factors, low regis- amplitude of skin conductance response than per-
tration (Factor 1) and sensation seeking (Factor 2) were sons with high thresholds (low registration and
consistent with the theory. Inconsistent with the theory sensation seeking).
was the finding that the items of sensory sensitivity and 2. Persons with high scores on sensation avoiding and
sensation avoiding loaded together on two factors (Factor 3 low registration would be quick to habituate but
and 4) (see Table 1). This finding is not entirely unexpect- for different reasons. (Sensation avoiding is a with-
ed because these two quadrants make up the low end of the drawal from the stimulus, whereas low registration
neurological threshold continuum. However, the finding has a limited response along with limited interest.)

78 January/February 2001, Volume 55, Number 1

Downloaded from http://ajot.aota.org on 09/24/2019 Terms of use: http://AOTA.org/terms
 Table 1
Four-Factor Model of the Adult Sensory Profile
Intended
Subscale Item Factor 1 Factor 2 Factor 3 Factor 4
A I avoid eye contact. .588
A I prefer sedentary activities. .584
S I become disoriented after bending over. .535
A I choose to spend my time in quiet activities. .509
R I have to ask people to repeat things. .503
R I don’t notice when my name is called. .494
R I seem slower than others when trying to follow an activity or task. .488
R I don’t notice when other people come in the room. .486
R I don’t pick up on what others are saying. .480
R It takes me more time to wake up in the morning. .454
A I keep the shades down. .454
R I am unaware of odors that others notice. .438
R I trip over or bump into things. .426
A I avoid situations where unexpected things might happen. .418
R I don’t seem to notice when someone touches me. .416
R I don’t seem to notice when my hands or face are dirty. .414
R I have a high pain tolerance. .411
S I gag easily with food textures or food utensils. .409
A I stay away from crowds. .408 .405
R I don’t get jokes as quickly as others. .408
R I miss street signs. .403
K I enjoy how it feels to move about. .674
K I like to wear colorful clothing. .582
K I choose to engage in physical activities. .573
K I do things on the spur of the moment. .569
K I find opportunities to visit places that have bright lights. .564
A I wear sunglasses when outside. .564
K I like to attend events with a lot of noise. .519
K When I see fresh flowers, I go over to smell them. .481
K I touch others when I’m talking. .457
K I find activities to perform in front of others. .444
K I like how it feels to get my hair cut. .440
K I enjoy being close to people who wear perfume or cologne. .439
K I work on two or more tasks at the same time. .438
K I like to go barefoot. .435
K I hum, whistle, sing, or make other noises. .420
S I am distracted if there is a lot of noise around. .743
A I stay away from noisy settings. .691
A When others are watching TV, I leave the room or ask them to turn it down. .635
S I find it difficult to work with background noise. .629
S I am bothered when I see lots of movement around me. .551
S I am bothered by unsteady or fast-moving images. .538
A I use strategies to drown out sound. .529
S I startle easily from unexpected or loud noises. .409
S I stay away from crowds. .408 .405
A I avoid stores with strong odors. .649
S I am uncomfortable in certain fabrics. .531
S I dislike having my back rubbed. .510
S I don’t like particular food textures. .573
S I don’t like strong-tasting mints or candies. .422
S I feel discomfort when brushing my teeth. .420
A I stay away from standing in line. .409
Note. Table includes items with factor loadings > .40. R = low registration; S = sensory sensitivity; A = sensation avoiding; K = sensation seeking.

3. Persons with high scores on sensory sensitivity and Physiological measure–skin conductance measures. Ampli-
sensation seeking would be slow to habituate but, tude and frequency of a skin conductance response is used
again, for different reasons. (Sensory sensitivity is a as a measure of attentional response allocation and process-
continued recognition of the stimulus, whereas ing of a stimulus (Dawson, Filion, & Schell, 1989). The
sensation seeking involves a desire to respond to auditory stimulus used consisted of 12, 108 dB white-noise
and an interest in the stimulus.) presentations of 1 sec duration. The interstimulus interval
was 30 sec to 50 sec. Electrodermal activity was recorded as
Participants. The total subscale scores from occupation- skin conductance responses from 8 mm silver chloride elec-
al therapy students were examined to identify the five high- trodes placed over the tips of the second and third fingers
est scoring students in each of the four quadrants. These of the participant’s nonpreferred hand. The skin conduc-
students were invited to complete the skin conductance pro- tance signal was recorded with a LabLinc-V1 system
tocol. All but one agreed to participate, and this student was 1
Coulbourn Instruments, 7462 Penn Drive, Allentown, Pennsylvania
replaced with the one with the next highest score. 18106.

The American Journal of Occupational Therapy 79

Downloaded from http://ajot.aota.org on 09/24/2019 Terms of use: http://AOTA.org/terms
 equipped with bioamplifier and isolated skin conductance responsivity to the stimulus with different habituation pat-
coupler. The signal was digitized at a rate of 20 Hz for 3000 terns, and the low registration group had a limited response
ms before to 8000 ms after the presentation of each stimu- and habituated quickly to the noise, both indicating a
lus. Skin conductance response amplitudes were computer behavioral response in accordance with a high neurological
scored off line as the change in conductance occurring threshold. The sensation seeking group also had a limited
within a latency window of 1 sec to 3 sec after stimulus response but did not habituate. This response is consistent
onset. Skin conductance level was scored at each 1-sec with a counteractive behavioral response to a high neuro-
interval across the recording window. logical threshold.
Skin conductance measures of responsiveness and tri-
als to habituation were obtained. Responsivity was mea- Step 4: Reliability of the Adult Sensory Profile—Revised
sured as the amplitude of the response to the first auditory Version With a Heterogeneous Group
stimulus. Trials to habituation was measured as the number Purpose. The purpose of this step was to determine whether
of stimuli presented before a participant had two consecu- the psychometric properties of the revised Adult Sensory
tive nonresponses to an orienting stimulus. Profile indicated improved reliability of the Sensation
Procedure. Written informed consent was obtained from Avoiding subscale, with continued adequate reliability of
all participants. Skin conductance measures were obtained in the Sensory Sensitivity, Low Registration, and Sensation
a laboratory located at a university setting. Participants were Seeking subscales.
instructed that they would hear a series of sounds and that Participants. The final version of the Adult Sensory
they should simply sit quietly during the process. Profile was administered to 93 adults for the purposes of
Analysis. One-way analyses of variance (ANOVAs; 1 x making group comparisons among persons with schizo-
4) were conducted to examine the differences among the phrenia, persons with bipolar disorder, and persons who
four groups on the skin conductance–dependent measures were mentally healthy. The group comparisons will be pre-
of responsivity and trials to habituation. Tukey’s test was sented in another article. The three groups were combined
used to examine pair-wise differences among the means for the current analyses to provide an adequate sample size.
(Green et al., 1997). Participants were recruited from consumers and staff at
Results. The results of the ANOVA were consistent three community mental health programs. Forty men and
with the a priori hypotheses. There was a significant differ- 53 women comprised the sample. The mean age was 38
ence in responsivity across the four groups: sensory sensi- years (SD = 11.2, range = 18–68 years).
tivity, sensation avoiding, low registration, and sensation Procedure. Written informed consent was obtained,
seeking, F(3, 17) = 8.38, p = .001. A Tukey’s test indicated and a research assistant administered the revised Adult
that the sensory sensitivity and sensation avoiding groups Sensory Profile to all participants.
were more responsive than the low registration and sensa- Analysis. As in the previous analysis, coefficient alpha for
tion seeking groups. There was also a significant difference each subscale and item correlations with each subscale using
in trials to habituation across the four groups, F(3, 17) = Pearson product–moment correlations were conducted.
46.85, p < .001. Habituation was defined as two consecu- Factor analysis was not used because of the small sample size.
tive nonresponses. A Tukey’s test indicated that the sensory Results. Item reliability was first examined by comput-
sensitivity and sensation seeking groups took more trials to ing coefficient alpha as an internal consistency estimate for
habituate than the sensation avoiding and low registration
each 15-item subscale. In this case, the Sensation Seeking
groups (see Table 2).
subscale had the poorest internal consistency (alpha = .60),
Examination of the two ANOVAs together revealed
whereas the Sensory Sensitivity (alpha = .78), Sensation
that each group was distinguished from all other groups by
Avoiding (alpha = .77), and Low Registration (alpha = .78)
a different pattern of responses consistent with the under-
subscales demonstrated strong internal consistency.
lying theory of Dunn’s Model of Sensory Processing. The
Compared with the previous analysis, the Sensation
groups with low neurological thresholds (sensory sensitivi-
Avoiding subscale improved, but the internal consistency
ty and sensation avoiding) responded intensely to stimuli
estimate for the Sensation Seeking subscale was worse. The
but had different habituation patterns. The sensory sensi-
items in the Sensation Seeking subscale remained virtually
tivity group responded intensely to the stimulus but did
not habituate to the noise, which is consistent with a Table 2
behavioral response in accordance with a low neurological Group Means for Skin Conductance Responsivity and
threshold. Conversely, the sensation avoiding group also Trials to Habituation
responded intensely to the stimulus but habituated quick- Responsivity Trials to Habituation
Group M (SD) M (SD)
ly to the noise, suggesting a behavioral response that coun-
Sensory sensitivity 2.04 (.20) 11.67 (0.82)
teracts the low neurological threshold. Sensation avoiding 2.05 (.17) 6.8 (1.09)
The two groups with a high neurological threshold Low registration 1.69 (.12) 6.2 (0.84)
Sensation seeking 1.63 (.19) 11.2 (1.09)
(low registration and sensation seeking) had restricted
80 January/February 2001, Volume 55, Number 1
Downloaded from http://ajot.aota.org on 09/24/2019 Terms of use: http://AOTA.org/terms
 the same for both phases of the study. Only one item everyday sensory stimuli.
received minor revision. Therefore, the decrement in inter- The item reliability measures and factor analysis sug-
nal consistency is probably due to the difference in the sec- gested that the Low Registration and Sensation Seeking
ond sample (i.e., smaller and including persons with severe subscales were internally consistent and could be distin-
mental illness). guished from the other quadrants. Conversely, the factor
Next, item reliability was examined using the correla- analysis revealed that the Sensory Sensitivity and Sensation
tion of each item on a subscale with the total score for each Avoiding subscales were not distinct quadrants because
Adult Sensory Profile subscale using Pearson product- items on these subscales loaded together. Other analyses
moment correlations. Only two items did not have their provided additional evidence that the Sensation Avoiding
highest correlation on the intended subscale. One sensation subscale was not discrete. The Sensation Avoiding subscale
avoiding item, “When I smell a strong odor in a store, I had a lower coefficient alpha than the other subscales, and
move to another section or leave the store,” had a higher 11 sensation avoiding items correlated higher with the total
correlation on the Sensory Sensitivity subscale (r = .299) scores on the other subscales than they did with their own
than on the Sensation Avoiding subscale (r = .275). A sen- total score. This evidence led to a substantial revision of the
sation seeking item, “I find opportunities to visit places Sensation Avoiding subscale to reflect more deliberate
that have bright lights and are colorful,” had a higher cor- avoidance behaviors.
relation on the Sensory Sensitivity subscale (r = .280) than To determine whether changes in the Sensation
on the Sensation Seeking subscale (r = .273). Overall, this Avoiding subscale were favorable, the performance of the
finding suggests a major improvement in the items because Adult Sensory Profile was reexamined with the sample of
before revision, 11 items on the Sensation Avoiding scale participants selected for group comparisons. Coefficient
had a higher correlation with an unintended subscale. alpha for the Sensation Avoiding subscale improved from
.66 with the original items to .77 with the revised items.
Discussion When item reliability was examined using the correlation
Psychometric Properties of the Adult Sensory Profile of each item on a subscale with a total subscale score, only
An examination of the reliability and validity of the Adult one item on the Sensation Avoiding subscale correlated
Sensory Profile guided changes in items and allowed for higher on other subscales than it did with the total score on
increased confidence in the instrument as a measure of sen- its own subscale. Overall, the Sensation Avoiding subscale
sory processing in adults. The remarkable agreement appears more distinct after item revision.
among the expert judges’ sorting of items according to the We acknowledge that skin conductance tests should be
four quadrants of Dunn’s Model of Sensory Processing readministered in conjunction with the revised Adult Sensory
indicated that the items could be categorized according to Profile to determine whether the same results are found with
the extant theory. An examination of the relationship the revised items. That study is currently under way.
between scores on the Adult Sensory Profile and skin con- Collectively, there is evidence to support the four sub-
ductance measures supported the construct validity of the scales of the Adult Sensory Profile as distinct constructs of
instrument. Responsivity and habituation to stimuli were sensory processing preferences. The evidence also indicates
examined for participants with high scores on each of the that the Sensation Avoiding subscale, which had the weak-
four quadrants. This analysis yielded unique patterns for est properties initially, was improved with item revision.
each group consistent with Dunn’s model. Participants
Clinical Applications of the Adult Sensory Profile
with high scores on sensory sensitivity were more respon-
sive and took longer to habituate. Participants with high The findings of these series of studies provide evidence that
scores on sensation avoiding were also more responsive but supports the reliability and validity for use of the Adult
quick to habituate. Participants with high scores on low Sensory Profile in practice settings. Information from the
registration were less responsive and quick to habituate, Adult Sensory Profile can have several intervention applica-
whereas participants with high scores on sensation seeking tions. First, providing the results to the person completing
were less responsive but slow to habituate. the measure gives insight into personal behavior and respons-
It should be noted that there is a discrepancy between es to different environments. Furthermore, this information
the sensation seeking results in this study and Zuckerman’s can increase understanding for family members, friends,
(1994) findings of rapid habituation among sensation seek- coworkers, and so forth regarding a person’s behaviors and
ers. The most likely explanation is related to a difference in responses to stimuli. The measure also can help explain areas
the sensation seeking construct of Zuckerman and the con- of conflict when persons have different sensory preferences.
struct of Dunn (1997). Zuckerman’s sensation seeking Service providers can use results of the Adult Sensory
emphasizes behaviors that involve impulsivity and risk tak- Profile to design more effective interventions. Such inter-
ing, whereas Dunn’s sensation seeking is manifested as ventions might include environmental adaptations to sup-
behaviors that indicate pleasure and interest in exposure to port performance. For example, persons with low registra-

The American Journal of Occupational Therapy 81

Downloaded from http://ajot.aota.org on 09/24/2019 Terms of use: http://AOTA.org/terms
 tion may need stimuli to be intensified, whereas persons method for applying a unique theoretical model toward the
with sensory sensitivity may need a reduction in intensity understanding of sensory processing. Although the data
or quantity of stimuli to avoid distractibility. Sensory pro- from this study initially support the reliability and validity
cessing preferences also are important when making judg- of the Adult Sensory Profile, further exploration of the
ments about environmental fit. For example, when a per- revised measure as well as studies with special populations
son is making decisions about a living situation or job will provide additional useful information regarding appli-
choice, results from the Adult Sensory Profile may enhance cation in clinical practice. ▲
the decision-making process. A workplace with quiet sur-
roundings and clear expectations might suit a sensation Acknowledgments
avoider, whereas a sensation seeker might prefer lots of We thank occupational therapy students Jason Wollenberg, Brant
Bermudez, and Jennifer Brunton for their assistance with data collection
activity and variety. There are also strategies that persons as well as all of the participants in the project. This study was complet-
can adopt when encountering adverse environments. For ed as partial fulfillment of the requirement for a Doctor of Philosophy
example, persons with sensory sensitivity may need to in Education Psychology and Research at the University of Kansas and
was funded by an AOTA/AOTF Dissertation Award.
develop strategies for maintaining focus in distracting envi-
ronments. References
American Occupational Therapy Association. (1994). Uniform
Future Directions for Research terminology for occupational therapy—Third edition. American Journal
of Occupational Therapy, 48, 1047–1054.
Efforts to ascertain reliability and validity of an instrument Baranek, G. T., Foster, L. G., & Berkson, G. (1997). Sensory
is a never-ending process. An important next step is to defensiveness in persons with developmental disabilities. Occupational
examine the psychometric properties of the Adult Sensory Therapy Journal of Research, 17, 173–185.
Crocker, L., & Algina, J. (1986). Introduction to classical and mod-
Profile as revised with a large normative population. ern test theory. Orlando, FL: Harcourt Brace Jovanovich College
However, too often psychometric properties of an instru- Publishers.
ment are obtained only on a normative population. The Dawson, M. E., Filion, D. L., & Schell, A. M. (1989). Is elicita-
reliability and validity of the instrument still needs to be tion of the autonomic orienting response associated with allocation of
processing resources? Psychophysiology, 26, 560–572.
examined for individuals from special populations. Dunn, W. (1994). Performance of typical children on the Sensory
Currently, studies are under way that are examining sensory Profile: An item analysis. American Journal of Occupational Therapy, 48,
processing patterns in schizophrenia, bipolar disorder, learn- 967–974.
ing disabilities, post-traumatic stress disorder, and aging. Dunn, W. (1997). The impact of sensory processing abilities on the
daily lives of young children and their families: A conceptual model.
This study supports the reliability and validity of the Infants and Young Children, 9, 23–35.
Adult Sensory Profile. The skin conductance data provide Dunn, W., & Brown, C. (1997). Factor analysis on the Sensory
physiological evidence to support construct validity for the Profile from a national sample of children without disabilities. American
Journal of Occupational Therapy, 51, 490–495.
instrument’s ability to measure four distinct patterns of
Falk-Kessler, J., Quittman, M. S., & Moore, R. (1988). The
sensory processing. The item revisions improved the inter- SCSIT: A potential tool for assessing neurological impairment in adult
nal consistency of the Sensation Avoiding subscale. A lin- psychiatric outpatients. Occupational Therapy Journal of Research, 8,
gering concern, however, is the reduction in internal con- 131–154.
Gorsuch, R. L. (1983). Factor analysis (2nd ed.). Hillsdale, NJ:
sistency on the Sensation Seeking subscale with the revised Erlbaum.
measure. Because the Sensation Seeking subscale received Green, S. B., Salkind, N. J., & Akey, T. M. (1997). Using SPSS for
very few changes, the lower estimate of the subscale’s inter- Windows: Analyzing and understanding data. Upper Saddle River, NJ:
nal consistency in Step 4 compared with Step 2 may be spe- Prentice Hall.
Kinnealey, M., Oliver, B., & Wilbarger, P. (1995). A phenomeno-
cific to the sample. The change in reliability from sample logical study of sensory defensiveness in adults. American Journal of
to sample further points to the importance of investigating Occupational Therapy, 49, 444–451.
the psychometric properties of the instrument in a variety Kohlmeyer, K. (1998). Evaluation of sensory and neuromuscular
of populations. performance components. In M. E. Neistadt & E. B. Crepeau (Eds.),
Willard and Spackman’s occupational therapy (9th ed., pp. 223–259).
Philadelphia: Lippincott-Raven.
Conclusion
Zuckerman, M. (1994). Behavioral expressions and biosocial bases of
Overall, the Adult Sensory Profile provides a credible sensation seeking. New York: Cambridge University Press.

82 January/February 2001, Volume 55, Number 1

Downloaded from http://ajot.aota.org on 09/24/2019 Terms of use: http://AOTA.org/terms