05 spm2 Manual CH Sample 041621 1

Selected material from the Sensory Processing Measure, Second Edition (SPMTM-2) manual.
Copyright © 2021 by Western Psychological Services (WPS®).

Provided by WPS for the sole purpose of introductory reference by qualified professionals. Not to be reprinted, excerpted, or distributed in whole
or in part without the prior written authorization of WPS ([email protected]). Full materials available for purchase at wpspublish.com.
5
Psychometric Properties
E
This chapter reviews reliability and validity evidence from the SPM‑2
PL
standardization and clinical validity studies. Validation of a scale is an
ongoing process, and it is hoped that the research presented here will
provide a foundation for further psychometric evaluations of the SPM‑2.
Reliability
M
Reliability refers to the consistency, stability, and consistency is a measure of the average intercorrela‑
predictability of test scores. The major purpose tions among the items that compose a scale. For the
SA
of any behavior rating scale is to facilitate clinical SPM‑2, Cronbach’s coefficient alpha (1988) was used
inferences in support of diagnostic and treat‑ to estimate internal consistency. Alpha takes the
ment decisions. If these inferences are to be valid, form of a correlation coefficient with a range of 0 to 1
they must be based on reliable scores. This sec‑ and represents the lower bound of a scale’s reliabil‑
tion describes seven approaches that were used ity. Tables 5.1 through 5.5 present alphas for each of
to estimate the reliability of the SPM‑2: internal the main SPM‑2 forms in the standardization sam‑
consistency, test–retest reliability, standard error of ple, subdivided by age. Table 5.6 presents alphas for
measurement, interrater reliability, alternate-forms the School Environment and Driving Environment
reliability, cross-rater concordance, and an online Forms. Table 5.7 presents alphas for all forms in the
and paper forms equivalence study. overall clinical sample except the Infant/Toddler age
level. Because the clinical sample size was low for
the Infant Form (n = 11), those cases were analyzed
Internal Consistency
with the Toddler Form (n = 43) by summing the item
Internal consistency represents the degree to which scores from the 46 items shared by both forms, thus
all items on a rating scale consistently measure the creating a provisional “total” score, which showed an
same dimension or trait. In statistical terms, internal alpha of .90.
SPM-2 • W-706M wpspublish.com
05_SPM_2_Manual_CH5_032421.indd 95 4/1/21 12:23 PM

In general, for behavior rating scales, alphas above results are consistent with other behavior rating
.80 are considered ideal, and above .70 are accept‑ scales of this kind and support a level of test–retest
able. Across the entire standardization sample, all reliability that is appropriate for clinical use of
of the Sensory Total (ST) scores were above .90. The the SPM‑2.
individual sensory scales showed a broader range
of coefficients: most were above .80, some were in
Standard Error of Measurement and
the .70s, and a few, mainly in the Infant/Toddler age
Confidence Intervals
level forms, were in the high .60s.
The standard error of measurement (SEM) statistic
These results were expected for two main reasons.
translates a reliability estimate into more practi-
First, aside from the ST, which is a 60‑item scale, the
cal terms by providing an index of how close an
SPM‑2 scales are quite brief at 10 items each. This
individual’s observed score is to their “true” score
brevity allows the SPM‑2 to measure its main con‑
(i.e., what the score would be if there were no mea‑
structs efficiently, but with somewhat lower internal
surement error). The SEM is inversely related to the
consistency than a longer scale. Second, with respect
reliability of the scale. Therefore, greater reliability
to the Infant/Toddler scales, relatively lower coeffi‑
means a smaller SEM, which increases the amount
cients are consistent with other measures of sensory
of confidence the clinician can have in the preci‑
processing in infants and toddlers, such as the Sen‑
sion of an observed test score. The SEM is calculated
sory Profile. This could be due to lesser stability in
using the formula SEM = SD √1−r , where SD is the
E
sensory responding among infants and toddlers, or
standard deviation of the scale, and r is the reliability
to the difficulty raters may have in correctly inter‑
coefficient of the scale. Tables 5.1 through 5.5 pres‑
preting their sensory processing.
PL ent SEM values and 95% confidence intervals for
Overall, these data suggest that the SPM‑2 scales the SPM‑2 scales in T‑score units, calculated sepa‑
show levels of internal consistency that support the rately for the test–retest and internal consistency
use of these scores in clinical applications. Further, reliability methods.
the SPM‑2 scales retain this level of internal consis‑
SEM values can be converted into confidence inter‑
tency in groups of individuals with different clinical
vals (CI) that give a range of possible values for the
diagnoses.
M
true score. For example, the 95% confidence inter‑
val represents the range of scores that has a 95%
Test–Retest Reliability probability of containing the true score. To obtain a
specific confidence interval for any observed SPM‑2
Test–retest reliability, also known as temporal sta-
score, add and subtract the tabled CI value from
SA
bility, refers to the stability of test scores over time.

the observed T‑score to create a score range. For
Test–retest reliability is estimated by having a single
instance, if a child’s observed T‑score on the ST scale
respondent rate a person using the same form on two
of the Preschool Home Form (Table 5.2) is 55, you
separate occasions and correlating the resulting scores.
would add and subtract 3.40, the test–retest reli‑
The SPM‑2 test–retest studies included participants ability CI value to get the confidence interval range.
from the standardization sample who were assessed In this example, there is a 95% probability that the
twice with the same form: an initial assessment and true T‑score lies in the range of 51.6 to 58.4. The test–
a second assessment approximately two weeks later. retest method is generally more relevant to practical
Overall, the demographic characteristics of the retest use of the SPM‑2 because it represents variability of
sample participants drawn from the standardiza‑ observed test scores over time. The alpha method, on
tion sample were similarly representative of the U.S. the other hand, represents the more abstract issue of
Census. Tables 5.1 through 5.5 present the test–retest the extent to which a scale measures a nonheteroge‑
correlations for each of the main forms across the neous construct. As noted previously, alpha provides
two-week retest interval. Most correlations were a lower bound on a scale’s reliability and thus is a
above .80, with a small number of correlations in the more conservative estimate, resulting in a wider, less
.70s, mostly at the Infant/Toddler age level. These precise, confidence interval.
96 SPM- 2 Chapter 5 Psychometric Properties

Table 5.1. SPM-2 Internal Consistency, Test–Retest Reliability, and Confidence Interval Estimates
in the Standardization Sample: Infant/Toddler Age Level Forms
Internal consistencya Test–retest reliabilityb

Age group (months)
No. of 4–6 7–9 4–9 95% CI 95% CI
Scale items α α α SEMc (+
−)
d r SEMc (+
−)
d
Infant Form
Vision (VIS) 10 .68 .77 .73 4.98 9.8 .72 3.65 7.15
Hearing (HEA) 10 .74 .71 .73 4.91 9.6 .74 3.63 7.11
Touch (TOU) 10 .68 .66 .67 5.47 10.7 .82 3.67 7.20
Taste and Smell (T&S) 10 .80 .76 .78 3.94 7.7 .71 4.38 8.58
Body Awareness (BOD) 10 .81 .72 .76 4.67 9.2 .77 4.02 7.88
Balance and Motion (BAL) 10 .77 .79 .78 4.37 8.6 .88 2.54 4.97
Sensory Total (ST) 60 .91 .91 .91 2.93 5.8 .91 2.36 4.62
Planning and Ideas (PLN) 10 .88 .81 .86 3.43 6.7 .77 3.85 7.54
Social Participation (SOC) 10 .87 .87 .87 3.40 6.7 .78 3.75 7.36
E
Age group (months)
No. of 10–20 21–30 10–30 95% CI 95% CI
−)
d r SEMc (+
−)
d
Toddler Form
Vision (VIS)
Hearing (HEA)
Touch (TOU)
10
10
10
PL .68
.74
.69
.68
.73
.71
.69
.73
.70
5.48
5.05
5.41
10.7
9.9
10.6
.78
.81
.79
4.11
3.98
3.86
8.05
7.80
7.57
M
Sensory Total (ST) 60 .92 .91 .92 2.79 5.5 .91 2.50 4.90
SA
Age group (months)

No. of 4–9 10–30 4–30 95% CI 95% CI
−)
d r SEMc (+
−)
d
Caregiver Self-Report Form

Vision (VIS) 10 .82 .77 .80 4.42 8.7 .85 3.52 6.91
Hearing (HEA) 10 .75 .77 .76 4.94 9.7 .84 3.68 7.21
Touch (TOU) 10 .72 .72 .72 5.20 10.2 .82 3.97 7.77
Sensory Total (ST) 60 .94 .93 .93 2.57 5.0 .93 2.30 4.50
Note. CI = confidence interval.

aCronbach’s alpha is given for Infant Form 4- to 6-month-old (n = 139) and 7- to 9-month-old (n = 127) strata, and 4- to 9-month-old sample (N = 266);
Toddler Form 10- to 20-month-old (n = 253) and 21- to 30-month-old (n = 168) strata, and 10- to 30-month-old sample (N = 421); and Caregivers of
4- to 9-month-olds (n = 363), 10- to 30-month-olds (n = 507), and the total sample (N = 870).
bTwo-week test–retest correlation; Infant Form n = 40; Toddler Form n = 91; Caregiver Self-Report Form n = 213.
cStandard error of measurement in T-score units, based on the overall sample for each form.
d95% confidence interval around the observed T-score, based on the overall sample for each form.
Reliability SPM- 2 97

in the Standardization Sample: Preschool Age Level Forms

Age group (years)
No. of 2–4 5 2–5 95% CI 95% CI
−)
d r SEMc (+
−)
d
Home Form
Vision (VIS) 10 .83 .89 .84 4.00 7.9 .84 3.39 6.64
Hearing (HEA) 10 .86 .91 .87 3.65 7.2 .93 2.40 4.71
Touch (TOU) 10 .79 .91 .81 4.36 8.6 .83 3.43 6.72
Sensory Total (ST) 60 .95 .98 .96 2.07 4.1 .96 1.73 3.40
E
School Form
Vision (VIS) 10 .87 .82 .87 3.42 6.7 .86 3.77 7.40
Hearing (HEA) 10 .88 .86 .88 3.21 6.3 .87 3.57 6.99
Touch (TOU)
Taste and Smell (T&S)
Body Awareness (BOD)
10
10
10
PL
.84
.88
.90
.79
.91
.90
.84
.88
.90
3.71
3.13
2.88
7.3
6.1
5.6
.83
.75
.87
3.89
4.45
3.07
7.63
8.72
6.02
Sensory Total (ST) 60 .97 .96 .97 1.73 3.4 .95 2.39 4.69
M

aCronbach’s alpha is given for Home Form 2- to 4-year-old (n = 445) and 5-year-old (n = 75) strata, and 2- to 5-year-old sample (N = 520); School
Form 2- to 4-year-old (n = 448) and 5-year-old (n = 91) strata, and 2- to 5-year-old sample (N = 539).
SA
bTwo-week test–retest correlation; Home Form n = 77; School Form n = 92.

in the Standardization Sample: Child Age Level Forms

Age group (years)
No. of 5–8 9–12 5–12 95% CI 95% CI
−)
d r SEMc (+
−)
d
Home Form
Vision (VIS) 10 .79 .81 .80 4.10 8.0 .85 3.45 6.76
Hearing (HEA) 10 .86 .87 .86 3.34 6.5 .83 3.61 7.08
Touch (TOU) 10 .71 .78 .74 4.74 9.3 .80 4.12 8.07
Sensory Total (ST) 60 .94 .95 .95 2.27 4.5 .79 4.22 8.28
E
School Form
Vision (VIS) 10 .83 .87 .85 3.75 7.4 .83 3.93 7.70
Hearing (HEA) 10 .84 .88 .86 3.57 7.0 .86 3.40 6.67
Touch (TOU)
10
10
10
PL .75
.85
.90
.87
.87
.90
.83
.85
.90
3.72
3.32
2.90
7.3
6.5
5.7
.88
.71
.87
2.97
4.68
3.53
5.83
9.17
6.91
Sensory Total (ST) 60 .96 .97 .96 1.73 3.4 .94 2.20 4.31
M

aCronbach’s alpha is given for Home Form 5- to 8-year-old (n = 475) and 9- to 12-year-old (n = 400) strata, and 5- to 12-year-old sample (N = 875);
School Form 5- to 8-year-old (n = 331) and 9- to 12-year-old (n = 224) strata, and 5- to 12-year-old sample (N = 555).
SA
bTwo-week test–retest correlation; Home Form n = 129; School Form n = 110.

in the Standardization Sample: Adolescent Age Level Forms

Age group (years)
No. of 12–13 14–15 16–17 18–21 12–21 95% CI 95% CI
Scale items α α α α α SEMc (+
−)
d r SEMc (+
−)
d
Home Form
Vision (VIS) 10 .84 .90 .86 .91 .88 3.13 6.1 .85 2.87 5.62
Hearing (HEA) 10 .87 .89 .88 .90 .88 3.22 6.3 .82 3.33 6.53
Touch (TOU) 10 .85 .86 .85 .89 .86 3.44 6.7 .87 3.05 5.98
Taste and Smell (T&S) 10 .87 .88 .86 .88 .87 3.30 6.5 .84 3.63 7.11
Body Awareness (BOD) 10 .89 .89 .86 .92 .89 3.02 5.9 .87 2.69 5.27
Balance and Motion (BAL) 10 .88 .92 .88 .92 .90 2.66 5.2 .75 3.39 6.65
Sensory Total (ST) 60 .96 .97 .97 .98 .97 1.70 3.3 .92 2.22 4.36
Planning and Ideas (PLN) 10 .91 .92 .90 .93 .91 2.76 5.4 .80 3.64 7.13
Social Participation (SOC) 10 .78 .81 .80 .82 .80 4.34 8.5 .89 2.90 5.68
E
School Form
Vision (VIS) 10 .90 .93 .90 .95 .92 2.28 4.5 .88 2.43 4.77
Hearing (HEA) 10 .90 .93 .90 .91 .91 2.41 4.7 .88 2.72 5.34
Touch (TOU)
10
10
10
.90
.92
.92
PL
.93
.93
.93
.90
.92
.93
.92
.94
.93
.91
.92
.92
2.22
2.20
2.06
4.4
4.3
4.0
.82
.87
.87
2.82
2.65
2.22
5.52
5.19
4.35
Sensory Total (ST) 60 .98 .99 .98 .99 .98 1.18 2.3 .94 2.14 4.19
M
Self-Report Form
Vision (VIS) 10 .83 .86 .83 .84 .84 3.82 7.5 .86 3.21 6.29
SA
Hearing (HEA) 10 .83 .86 .82 .83 .84 3.89 7.6 .87 3.42 6.71
Touch (TOU) 10 .78 .79 .70 .83 .77 4.56 8.9 .87 3.14 6.16
Taste and Smell (T&S) 10 .83 .83 .83 .88 .84 3.86 7.6 .82 4.02 7.88
Body Awareness (BOD) 10 .83 .87 .84 .85 .85 3.69 7.2 .85 3.08 6.04
Sensory Total (ST) 60 .95 .96 .95 .96 .96 2.03 4.0 .93 2.40 4.71

aCronbach’s alpha is given for Home Form 12- to 13-year-old (n = 288), 14- to 15-year-old (n = 319), 16- to 17-year-old (n = 327), and 18- to 21-year-old
(n = 83) strata, and 12- to 21-year-old sample (N = 1,021); School Form 12- to 13-year-old (n = 202), 14- to 15-year-old (n = 190), 16- to 17-year-old
(n = 199), and 18- to 21-year-old (n = 54) strata, and 12- to 21-year-old sample (N = 645); and Self-Report Form 12- to 13-year-old (n = 258),
14- to 15-year-old (n = 256), 16- to 17-year-old (n = 314), and 18- to 21-year-old (n = 129) strata, and 12- to 21-year-old sample (N = 957).
bTwo-week test–retest correlation; Home Form n = 114; School Form n = 87; Self-Report Form n = 104.

in the Standardization Sample: Adult Age Level Forms
SPM-2 • W-706M
Reliability
Age group (years)
05_SPM_2_Manual_CH5_032421.indd 101
No. of 21–30 31–40 41–50 51–64 65–87 21–87 95% CI 95% CI
d d
Scale items α α α α α α SEMc (+
−) r SEMc (+
−)
Self-Report Form
Vision (VIS) 10 .78 .75 .87 .84 .82 .81 4.35 8.5 .86 3.45 6.76
Hearing (HEA) 10 .82 .84 .85 .87 .86 .84 3.97 7.8 .89 3.18 6.24
Touch (TOU) 10 .74 .75 .80 .83 .77 .78 4.73 9.3 .80 4.02 7.89
Taste and Smell (T&S) 10 .78 .71 .80 .80 .79 .78 4.69 9.2 .83 4.14 8.11
Body Awareness (BOD) 10 .81 .84 .90 .86 .89 .86 3.74 7.3 .83 3.86 7.57
Balance and Motion (BAL) 10 .82 .75 .85 .86 .84 .82 4.24 8.3 .83 3.42 6.70
Sensory Total (ST) 60 .94 .93 .96 .95 .95 .95 2.28 4.5 .93 2.38 4.66
SA
Planning and Ideas (PLN) 10 .86 .84 .86 .85 .87 .86 3.79 7.4 .78 4.12 8.07
Social Participation (SOC) 10 .75 .73 .73 .76 .82 .75 5.01 9.8 .90 2.38 4.66
Rater Report Form

M
Vision (VIS) 10 .83 .82 .90 .88 .91 .86 3.41 6.7 .79 4.17 8.17
Hearing (HEA) 10 .84 .83 .86 .88 .88 .85 3.69 7.2 .79 4.08 8.00
Touch (TOU) 10 .76 .62 .83 .88 .85 .80 4.25 8.3 .76 4.12 8.08
Taste and Smell (T&S) 10 .86 .84 .89 .87 .89 .87 3.44 6.7 .84 3.58 7.02
Body Awareness (BOD) 10 .83 .74 .89 .90 .90 .86 3.36 6.6 .84 3.21 6.30
Balance and Motion (BAL) 10 .78 .66 .89 .90 .89 .86 3.44 6.7 .80 3.41 6.69
PL
Sensory Total (ST) 60 .94 .92 .97 .98 .97 .96 1.98 3.9 .90 2.80 5.49
Planning and Ideas (PLN) 10 .86 .84 .86 .90 .90 .87 3.36 6.6 .79 3.75 7.36
Social Participation (SOC) 10 .82 .80 .81 .84 .89 .82 4.05 7.9 .77 4.77 9.36
E
aCronbach’s alpha is given for Self-Report Form 21- to 30-year-old (n = 184), 31- to 40-year-old (n = 198), 41- to 50-year-old (n = 143), 51- to 64-year-old (n = 111), and 65- to 87-year-old (n = 75) strata, and
21- to 87-year-old sample (N = 711); and Rater Report Form 21- to 30-year-old (n = 185), 31- to 40-year-old (n = 181), 41- to 50-year-old (n = 106), 51- to 64-year-old (n = 81), and 65- to 87-year-old (n = 51)
strata, and 21- to 87-year-old sample (N = 604).
bTwo-week test–retest correlation; Self-Report Form n = 52; Rater Report Form n = 58.
SPM- 2 101
wpspublish.com
4/1/21 12:23 PM
Table 5.6. SPM-2 Internal Consistency Estimates:
Child School Environment and Adolescent and Adult Driving Environment Forms
Internal
Scale No. of items n consistencya
Child School Environment Forms

Art (ART) 15 171 .93
School Bus (BUS) 15 170 .91
Cafeteria (CAF) 15 208 .89
Music (MUS) 15 106 .92
Physical Education (PHY) 15 173 .88
Recess/Playground (REC) 15 179 .93
Adolescent and Adult Driving Environment Forms

Driving Self-Report 18 777 .91
Driving Rater Report 18 292 .89
aCronbach’s alpha.
E
Table 5.7. SPM-2 Internal Consistencya Estimates:
Mixed Clinical Sample
PL Age level
Preschool Child Adolescent Adult
Self- Self- Rater
Home School Home School Home School Report Report Report
Scale Form Form Form Form Form Form Form Form Form
Vision (VIS) .86 .88 .84 .88 .86 .88 .85 .87 .89
Hearing (HEA) .87 .85 .88 .83 .81 .89 .81 .89 .89
M
Touch (TOU) .79 .77 .77 .76 .82 .82 .83 .74 .74
Taste and Smell (T&S) .85 .84 .82 .85 .86 .80 .89 .77 .88
Body Awareness (BOD) .91 .94 .85 .89 .89 .82 .92 .88 .89
SA
Balance and Motion (BAL) .87 .85 .79 .81 .90 .85 .90 .80 .88
Sensory Total (ST) .95 .96 .95 .95 .96 .96 .96 .95 .96
Planning and Ideas (PLN) .92 .94 .92 .94 .93 .92 .88 .91 .93
Social Participation (SOC) .93 .94 .92 .94 .82 .92 .74 .80 .90
Note. Preschool N = 107; Child N = 218; Adolescent N = 105; Adult N = 120. All scales are 10 items, except ST (60 items). Infant/Toddler coefficients
(N = 97) are described in the main text.
aCronbach’s alpha.

Interrater Reliability assessed with the Toddler Form and Preschool
Home Form; five-year-olds were assessed with the
Interrater reliability refers to the consistency of Preschool and Child Home Forms; 12-year-olds
scores obtained from different respondents (such as were assessed with the Child and Adolescent Home
two parents) who rated the same person using the Forms; and 21-year-olds were assessed with the
same form. The SPM‑2 interrater reliability study Adolescent and Adult Self‑Report Forms. In addition,
consisted of 188 participants (infant through adult) five-year-olds were assessed with the Preschool and
from the standardization sample who were assessed Child School Forms, and 12-year-olds were assessed
twice with the same form: once by the primary with the Child and Adolescent School Forms.
respondent, and once by a second respondent. Spe‑
cifically, infants and toddlers were assessed by two Alternate-forms studies emphasize the equivalency
parents or other caregivers; preschoolers, children, of parallel forms intended for different age ranges.
and adolescents were assessed by two parents or As expected, the pairs of parallel forms across all age
other caregivers with the Home Form, or by two levels showed a strong median correlation of .69 for
teachers or other school staff with the School Form; the 10‑item scales, and .84 for the 60‑item Sensory
and adults were assessed by two raters with the Rater Total. These studies suggest that, in clinical applica‑
Report Form. tions, use of either the “younger” or “older” forms
with individuals whose ages are within the overlap‑
In comparison to internal consistency and test–retest ping ranges of different form levels will produce
E
reliability, interrater reliability is subject to greater roughly equivalent results. Therefore, determining
measurement error, due to the two respondents’ which form to use when an examinee’s age falls
differing experiences with the rated individual. within the range of two forms is primarily a clinical
Therefore, scores from the two respondents are
PL decision. For more information about selecting the
expected to correlate moderately with each other, appropriate form, see Chapter 2, “Administration
and less strongly than internal consistency or test– Instructions by Age Level.”
retest correlation. A clinical corollary of this effect
is the importance of obtaining multiple ratings on
individuals, to incorporate the unique perspective of Cross-Rater Concordance
each rater in the assessment process.
M
Cross-rater concordance, also known as cross-form
As expected, the correlations were moderate across consistency, refers to the correlation between scores
all age levels, with a median correlation of .63 for the obtained from two respondents rating the same
10‑item scales and .78 for the 60‑item Sensory Total. person on two different forms. By convention, these
studies are usually reported alongside traditional
SA
These results are consistent with other behavior

rating scales of this kind and support a level of inter‑ reliability studies. However, in contrast to interrater
rater reliability that is appropriate for clinical use of and alternate-forms reliability, cross-rater consis‑
the SPM‑2. tency is closer to a metric of validity. This is because
it uses both different raters and different forms. In
the case of the SPM‑2, for example, ratings are given
Alternate-Forms Reliability not only by different raters, but by raters who have
Alternate-forms reliability refers to the consistency of different role relationships to the person, such as
scores obtained from the same respondent rating the parents and teachers. Parents and teachers observe
same person on two different forms. For the SPM‑2, children in different settings, during different time
these studies were conducted with individuals whose periods, and with different background knowledge
ages were within the overlapping year between two and expectations. In addition, the forms themselves
age levels (such as a 5-year-old who was rated with are designed to measure the SPM‑2 constructs in
both the Preschool and Child Home Forms). This different environments, where sensory process‑
study consisted of 54 participants (infant through ing behaviors are likely to vary based on different
adult) from the standardization sample who were stimuli and conditions. Therefore, the cross-rater
assessed with each alternate form. Specifically, concordance correlations (different raters, differ‑
nine- and ten-month-olds were assessed with both ent forms) were expected to be lower than both the
the Infant and Toddler Forms; two-year-olds were interrater study (different raters, same form) and the

alternate-forms study (same rater, parallel forms), These findings attest to the utility and clinical value
and also lower than the interscale correlations (same of having separate forms and raters for the home
rater, different traits). and school environments, as well as for self‑report
and rater report forms. Thus, in clinical applications,
This cross-rater concordance study consisted of
SPM‑2 users should strive to obtain multiple ratings
2,915 participants (infant through adult) from the
on the same individual in order to gain insight into
standardization sample who were assessed with at
the individual’s functioning in different settings, as
least two forms in their age level. The Preschool and
perceived by different raters.
Child age levels included Home and School com‑
parisons; the Adolescent age level included Home
and School, Home and Self‑Report, and School and Online and Paper Forms Equivalence
Self‑Report comparisons; and the Adult age level Study
included Self‑Report and Rater Report comparisons.
An equivalence study examined the extent to
Infant/Toddler participants were excluded because
which SPM-2 online and paper administrations,
only one SPM‑2 form is used to assess an infant
the formats typically used for remote and in-person
or toddler.
assessments, show equivalent results. The SPM-2
As expected, the median cross-rater concordance online and paper forms equivalence study consisted
correlations were lower than the interrater (.63 of 173 participants from the standardization sample
E
and .78) and alternate-forms (.69 and .84) correla‑ who completed two research forms in the same
tions described earlier in this chapter, as well as testing session: one online form (typically used for
the median interscale correlations (.62 and .79) remote assessments), and one paper form (typically
described in the Validity section of this chapter.
PL used for in-person assessments). Across all age levels,
Across all age levels, the cross-rater concordance median correlations were high: .92 for the Sensory
form pairs showed a median correlation of .47 for the Total (ST), and between .81 and .87 for the other eight
10‑item scales, and .59 for the 60‑item Sensory Total. scales. These findings suggest that online and print
This suggests that differences between observers or (and, by extension, remote and in-person) admin‑
environments affect SPM‑2 ratings more strongly istrations can be treated as equivalent methods for
than do differences among the sensory systems and administering the SPM-2.
M
the constructs of praxis and social participation.
SA

Validity
The concept of test validity has both theoretical and Construct Validity
practical dimensions. The theoretical aspects of
Construct validity refers to how well a test performs
validity deal with whether a scale measures the
in measuring a theoretical construct of interest. For
constructs that it purports to measure. The practical
the SPM‑2, Ayres Sensory Integration theory defines
components of validity refer to what sorts of clinically
the primary constructs of interest: processing in the
relevant information can be inferred from test scores.
visual, auditory, tactile, olfactory/gustatory, proprio‑
To address these issues, this section presents evidence
ceptive, and vestibular systems, along with praxis
regarding the content validity, construct validity, and
(motor planning and ideation) and social participa‑
criterion-related validity (i.e., the validity evidence
tion. The SPM‑2 is designed to provide access to these
based on clinical groups) of the SPM‑2 scales.
theoretical constructs through its T‑scores. There‑
fore, the construct validity of the SPM‑2 is based on
Content Validity its content validity, already described, as well as its
structural validity and nomological validity, both of
Content validity refers to the extent to which a
which are described in the following sections.
test represents all facets of a given construct. The
E
facet structure can be described as a matrix of the
content domains and dimensions that define a Structural Validity
construct according to theory. For the SPM‑2, the
PL Structural validity refers to the organization of
sensory domains (e.g., vision) combined with the
SPM‑2 items into scales that purport to measure vari‑
vulnerability dimensions (e.g., over-reactivity),
ous aspects of the constructs of sensory integration
when represented across all age levels and settings,
and processing. Evidence of the structural validity
comprise the facet structure of the SPM‑2. Content
of the SPM‑2 is demonstrated through its interscale
validity is built into a behavior rating scale by care‑
correlations and its factor structure.
fully describing the facets to be sampled, writing
M
items that represent those facets, subjecting the items
Interscale Correlations
to expert scrutiny, and then analyzing the result‑
ing item pool to ensure that the collection of items Tables 5.8 through 5.12 show the correlations among
retained in the final scales represents the full facet the SPM‑2 scales in the standardization sample. It
structure of the construct while retaining desirable is instructive to compare these correlations to the
SA
psychometric characteristics. internal consistency (alpha) coefficients in Tables 5.1

through 5.5. Across all SPM‑2 forms, the internal
Like its predecessors, the SPM‑2 is anchored in Ayres consistency reliability coefficients are higher than
Sensory Integration theory. This theory proposes that the interscale correlations. This indicates that each
the integration and processing of sensory inputs is a scale’s items are more strongly related to each other
critical neurobehavioral process that strongly affects than they are to other scales, which supports the sep‑
development. The theory holds that those with arate scoring and interpretation of the SPM‑2 scales,
compromised sensory processing may be unable to and by implication supports the structural validity
learn efficiently, regulate their emotions, or function of the instrument. When comparing the alphas to
at an expected level in daily activities. Difficulties the interscale correlations, however, it is impor‑
at the level of sensory integration and processing tant to exclude any scales that share items, because
often contribute to impairment in complex, higher these common items inflate interscale correlations.
level functions, such as praxis (the ability to plan and For that reason, the Sensory Total (ST) correlations
organize movement) and social participation. The should not be included in this type of comparison.
development procedures described in Chapter 4
generate confidence in the content validity of the
SPM‑2, ensuring that the items in each form rep‑
resent sensory integration and processing across
relevant content facets.
Validity SPM- 2 105

Table 5.8. SPM-2 Interscale Correlations in the Standardization Sample:
Infant/Toddler Age Level
Interscale correlationsa
Scale VIS HEA TOU T&S BOD BAL ST PLN SOC
Infant Form
Vision (VIS) — — — — — — — — —
Hearing (HEA) .34 — — — — — — — —
Touch (TOU) .36 .46 — — — — — — —
Taste and Smell (T&S) .25 .40 .44 — — — — — —
Body Awareness (BOD) .34 .33 .50 .29 — — — — —
Balance and Motion (BAL) .22 .53 .57 .31 .39 — — — —
Sensory Total (ST) .67 .70 .77 .58 .67 .68 — — —
Planning and Ideas (PLN) .09 .41 .26 .20 .24 .39 .35 — —
Social Participation (SOC) −.14 .37 .17 .06 −.03 .32 .14 .40 —
Toddler Form
Vision (VIS) — — — — — — — — —
E
Hearing (HEA) .67 — — — — — — — —
Touch (TOU) .48 .62 — — — — — — —
Taste and Smell (T&S) .33 .41
PL .51 — — — — — —
Sensory Total (ST) .76 .83 .80 .64 .80 .75 — — —
Social Participation (SOC) .20 .67 .25 .29 .31 .38 .37 .40 —
M
Caregiver Self-Report Form
Vision (VIS) — — — — — — — — —
Hearing (HEA) .67 — — — — — — — —
SA
Touch (TOU) .57 .63 — — — — — — —

Sensory Total (ST) .82 .84 .81 .80 .76 .80 — — —
Note. Infant Form N = 266; Toddler Form N = 421; Caregiver Self-Report Form N = 870. Because the Infant/Toddler participants were the primary
determinants of the demographic groups, the Caregiver sample was not desampled and thus shows a larger sample size than the Infant/Toddler
sample described in Table 4.1.
aCorrected Pearson correlation coefficients.

Preschool Age Level
Home Form
Vision (VIS) — — — — — — — — —
Hearing (HEA) .70 — — — — — — — —
Touch (TOU) .64 .68 — — — — — — —
Sensory Total (ST) .85 .86 .85 .76 .83 .76 — — —
School Form
Vision (VIS) — — — — — — — — —
E
Hearing (HEA) .77 — — — — — — — —
Touch (TOU) .68 .77 — — — — — — —
PL .64 — — — — — —
Sensory Total (ST) .85 .88 .88 .70 .79 .78 — — —
M
Note. Home Form N = 520; School Form N = 539.
SA
Validity SPM- 2 107

Child Age Level
Home Form
Vision (VIS) — — — — — — — — —
Hearing (HEA) .62 — — — — — — — —
Touch (TOU) .62 .64 — — — — — — —
Sensory Total (ST) .78 .79 .84 .78 .82 .77 — — —
School Form
Vision (VIS) — — — — — — — — —
E
Hearing (HEA) .73 — — — — — — — —
Touch (TOU) .66 .71 — — — — — — —
PL .60 — — — — — —
Sensory Total (ST) .87 .87 .85 .69 .86 .83 — — —
M
Note. Home Form N = 875; School Form N = 555.
SA

Adolescent Age Level
Home Form
Vision (VIS) — — — — — — — — —
Hearing (HEA) .74 — — — — — — — —
Touch (TOU) .66 .65 — — — — — — —
Sensory Total (ST) .83 .82 .82 .79 .80 .72 — — —
School Form
Vision (VIS) — — — — — — — — —
E
Hearing (HEA) .72 — — — — — — — —
Touch (TOU) .67 .68 — — — — — — —
Taste and Smell (T&S) .58 PL .68 .62 — — — — — —
Sensory Total (ST) .81 .87 .77 .77 .78 .78 — — —
M
Self-Report Form
Vision (VIS) — — — — — — — — —
Hearing (HEA) .77 — — — — — — — —
SA
Touch (TOU) .70 .74 — — — — — — —

Sensory Total (ST) .86 .89 .87 .82 .84 .84 — — —
Note. Home Form N = 1,014; School Form N = 645; Self-Report Form N = 957.
Validity SPM- 2 109

Adult Age Level
Self-Report Form
Vision (VIS) — — — — — — — — —
Hearing (HEA) .71 — — — — — — — —
Touch (TOU) .62 .61 — — — — — — —
Sensory Total (ST) .84 .82 .83 .76 .74 .82 — — —
Rater Report Form

Vision (VIS) — — — — — — — — —
E
Hearing (HEA) .71 — — — — — — — —
Touch (TOU) .70 .72 — — — — — — —
PL .65 — — — — — —
Sensory Total (ST) .84 .86 .87 .81 .72 .76 — — —
M
Note. Self-Report Form N = 711; Rater Report Form N = 604.
SA

Factor Analysis samples. A separate chi-square difference test was
Factor analysis begins with a correlation matrix performed to examine the efficacy of this multifactor
encompassing a large set of item responses, and then model over a general one-factor model. Across all
reduces that multitude of interrelationships to a samples, the eight-factor model significantly improves
much smaller set of underlying (or latent) variables. model fit compared to the one-factor model:
Because these latent variables often represent usable
· I/T Infant: x2(28) = 60.42, p < .001
scales, factorial validity, or the extent to which the
theoretical structure of a scale is recoverable in a set · I/T Toddler: x2(28) = 773.47, p < .001
of test scores, is an essential part of establishing a
test’s structural validity. The factorial validity of the
· I/T Caregiver Self-Report: x2(28) = 1,711.66,
p < .001
SPM‑2 is based in part on prior analyses of the SPM,
· Preschool Home: x2(28) = 1,232.48, p < .001
·
the SPM‑P, and their predecessors.
Preschool School: x2(28) = 1,070.60, p < .001
The theoretical foundations of the SPM‑2 assert that
its eight domains constitute important aspects of · Child Home: x2(28) = 1,846.55, p < .001
sensory processing, praxis, and social participation.
· Child School: x2(28) = 988.23, p < .001
In addition, Ayres (1972, 1979, 2005) postulated
· Adolescent Home: x2(28) = 1,555.59, p < .001
·
a further set of theoretical constructs related to
sensory integration vulnerabilities (e.g., under- and Adolescent School: x2(28) = 1,374.85, p < .001
E
over-reactivity to stimuli, sensation-seeking behav‑
· Adolescent Self‑Report: x2(28) = 1,121.36, p < .001
ior, perception difficulties, etc.). In both the SPM and
the SPM‑P, as well as their precursors, factor analyses · Adult Rater Report: x2(28) = 1,130.22, p < .001
PL
showed a better fit to a sensory systems factor model
than a vulnerabilities factor model.
· Adult Self‑Report: x2(28) = 1,314.77, p < .001
These findings support a suitable fit of the predicted
Based on these conceptual and practical consid‑ multifactor model such that the items comprising
erations, the factor structure of the SPM‑2 was the SPM‑2 meaningfully distinguish distinct aspects
examined using an eight-factor theoretical model. of sensory processing.
Confirmatory factor analysis was applied to evalu‑
M
To summarize, prior evidence from the SPM and
ate the model, with modification if necessary, and
the SPM-P, combined with the current examinations
analyzed using Mplus (Version 7) software (Muthén
of cross-rater concordance, interscale correlations,
& Muthén, 2012). In the hypothesized model, items
and confirmatory factor analyses, provides strong
were identified as reflecting Vision, Hearing, Touch,
SA
and consistent support for separate scoring and

Taste and Smell, Body Awareness, Balance and
interpretation of the Vision, Hearing, Touch, Taste
Motion, Planning and Ideas, and Social Participation.
and Smell, Body Awareness, Balance and Motion,
The model fit statistics for an eight-factor model Planning and Ideas, and Social Participation scales.
are presented in Table 5.13 for the standardization
Validity SPM- 2 111

Table 5.13. SPM-2 CFA Fit Statistics for Eight-Factor Model Across Standardization Samples
chi-
Standardization sample n square df p WRMR RMSEA CFI TLI
Infant/Toddler Forms
Infant 266 4,355.45 3,052 <.001 1.455 .040 .814 .807
Toddler 421 5,031.87 3,052 <.001 1.526 .039 .825 .819
Caregiver Self-Report 870 7,212.06 3,052 <.001 1.829 .040 .866 .861
Preschool Forms
Home 520 5,112.48 3,052 <.001 1.432 .036 .926 .923
School 539 5,672.46 3,052 <.001 1.626 .040 .931 .929
Child Forms
Home 875 5,389.39 3,052 <.001 1.453 .030 .943 .941
School 555 4,414.39 3,052 <.001 1.340 .028 .957 .956
Adolescent Forms
Home 1,014 7,547.40 3,052 <.001 1.706 .038 .922 .919
E
School 645 4,495.00 3,052 <.001 1.279 .027 .972 .971
Self-Report 957 6,654.25 3,052 <.001 1.570 .035 .930 .928
Adult Forms
Self-Report
Rater Report
711
604
6,310.52
4,916.25
PL 3,052
3,052
<.001
<.001
1.599
1.330
.039
.032
.893
.932
.889
.930
Note. The estimator WLSMV was used for analysis, as it is more appropriate for categorical data. In such cases, the chi-square value for categorical
data cannot be used for chi-square difference because the chi-square is adjusted to obtain an accurate p-value. A separate chi-square difference test
was performed for each sample to examine the efficacy of the multifactor model over the single-factor model. These results are reported within the
body of the text.
CFA = confirmatory factor analysis; df = degrees of freedom; p = the probability, testing against the null hypothesis, that the population RMSEA is
M
zero; WRMR (weighted root-mean-square residual) = average correlation residuals; RMSEA (root-mean-square error of approximation) = function
of chi-square test of close fit; CFI (Bentler comparative fit index) = measure of proportionate improvement in model fit; TLI (Tucker-Lewis index) =
measure of proportionate improvement in model fit.
SA

Nomological Validity forms in the same testing session. The equivalence
study was primarily conducted in the clinical
The SPM‑2 includes new studies of nomological
sample, with a smaller sample drawn from the
validity, or the degree to which a test conforms to
standardization sample mainly for the purpose of
a set of expected relationships with other variables
comparison. Separate studies were conducted for
(see Cronbach & Meehl, 1955). The SPM‑2 nomo‑
each of the forms, and the order of administration
logical validity studies were based on concurrent
was counterbalanced across each data collection site.
administration of the SPM‑2 and relevant measures
to assess convergent and divergent relationships, The results of these studies are presented in
or the degree of association between the SPM‑2 and Tables 5.14 and 5.15. The correlation coefficients,
measures of similar and dissimilar constructs. with a median correlation of .80, show a moderate
to strong degree of equivalence between versions.
Equivalence With the SPM and SPM‑P In general, equivalence is affected by several factors.
A series of analyses examined the level of equiva‑ Stronger equivalence is driven by the high degree of
lence between scores from the SPM‑2 Child and similarity in item content between the two versions,
Preschool age level forms and their predecessors, and weaker equivalence by revisions to the compo‑
the analogous forms of the Sensory Processing sition of the scales. In this case, because over 70%
Measure (SPM) and the Sensory Processing Measure– of the item content was the same, and many of the
Preschool (SPM‑P). Participants in the standardiza‑ changes in the other 30% were edits to existing items
E
tion and clinical samples rated the same individuals rather than new items, these equivalence studies
on the SPM‑2 and the corresponding SPM/SPM‑P produced the expected results.
PL
Table 5.14. Correlations Between SPM-2 and SPM-P T-Scores:
Preschool Age Level
Standardization sample Clinical sample

Corrected Corrected
SPM-P n r n r
Home Form 25 64
M
Vision (VIS) .69 .80
Hearing (HEA) .81 .80
Touch (TOU) .82 .74
SA
Body Awareness (BOD) .42 .84

Balance and Motion (BAL) .81 .85
Sensory Total (ST) .85 .95
Planning and Ideas (PLN)a .80 .86
Social Participation (SOC) .68 .87
School Form 32 73
Touch (TOU) .43 .50
Note. Standard scores for the T&S scale were not included in the SPM-P.
aThe Planning and Ideas (PLN) scale was abbreviated PLA in the SPM-P.
Validity SPM- 2 113

Table 5.15. Correlations Between SPM-2 and SPM T-Scores:
Child Age Level

Corrected Corrected
SPM n r n r
Home Form 13 149

Touch (TOU) .93 .75
School Form 14 113

E
Touch (TOU) .93 .70
Balance and Motion (BAL)
Sensory Total (ST)
Planning and Ideas (PLN) a
PL .78
.96
.86
.74
.90
.81
Note. Standard scores for the T&S scale were not included in the SPM.
aThe Planning and Ideas (PLN) scale was abbreviated PLA in the SPM.
M
SA

Convergence With the Sensory Profile stronger, reaching into the .60s and .70s. The Child
For the SPM‑2, the chief convergent measures age level (Table 5.17) showed a similar pattern of
of interest are the Sensory Profile forms, which correlations, but generally stronger, with all of the
included the Sensory Profile 2 (SP-2; Dunn, 2014) SPM‑2 Home and School scales showing correlations
and the Adolescent/Adult Sensory Profile (SP; Brown in the .50–.70 range with their corresponding SP‑2
& Dunn, 2002). Accordingly, convergent validity scales, with the exception of Social Participation on
studies were conducted at each age level. In particu‑ the Home Form (.41).
lar, six SPM-2 forms (Infant and Toddler, Preschool At the Adult and Adolescent age levels, the results
Home and School, Child Home and School) were were more variable (Table 5.18). These studies were
compared to comparable forms of the Sensory conducted using the Adolescent/Adult Sensory Pro‑
Profile 2. Two SPM‑2 forms (Adolescent and Adult file. For the most part, the SP scales, which are more
Self-Reports) were compared to the Adolescent/Adult limited in scope than those of the SP‑2 and only
Sensory Profile because the SP‑2 does not include include self-report forms, correlated less strongly
adolescent or adult forms. The remaining SPM‑2 and consistently with the SPM‑2 scales in com‑
forms (Adult Rater Report and I/T Caregiver Self- parison to the SP‑2. In particular, while the overall
Report) did not have a comparable Sensory Profile pattern of significant results was reflective of the
form, and so were not included in these studies. shared constructs, most of the SP Adolescent/Adult
These studies were conducted in the clinical sample, correlations were nonsignificant with the SPM‑2
E
with subsamples showing demographic characteris‑ Adolescent Self-Report Form. The Adult age level was
tics similar to those of the clinical sample as a whole more robust, showing a pattern of correlations simi‑
(see Table 4.13). Tables 5.16 through 5.18 show the lar to the lower age levels, with the main difference
PL
correlation matrices between the SPM‑2 and the
SP/SP‑2 forms. Correlations which were signifi‑
cant at the p < .01 level are shown in bold italics. As
expected, scales with similar content showed stron‑
being a broader range of significant correlations
among similar scales (.32 to .80).
Overall, these results show an expected relation‑
ger correlations. ship between scores from two measures of similar
constructs. In addition to the expected pattern of cor‑
Because many of the sensory domains of the SP‑2 relations among the sensory scales, supplementary
M
cannot be scored at the youngest ages due to low reli‑ analyses revealed a pattern of moderate correlations
abilities (Dunn, 2014), the most relevant correlations (.30 to .70) between the Preschool, Child, Adolescent,
at the Infant/Toddler age level were between the SP‑2 and Adult age level SPM‑2 scales and the SP/SP‑2
general processing score and the SPM‑2 scales. These vulnerability quadrants. These quadrants (seeking,
SA
correlations ranged from .44 to .61 across the SPM‑2 avoiding, sensitivity, and registration) correspond
scales. As expected, the SPM‑2 Sensory Total showed roughly to the main SPM‑2 sensory vulnerabil‑
the highest correlation (.63). ity dimensions (over-reactivity, under-reactivity,
At the Preschool age level (Table 5.16), both the sensory seeking, etc.) underlying the SPM‑2 items
SPM‑2 Home and School Forms showed a pattern across all forms. These results, taken together, sug‑
of expected correlations with the SP‑2 forms. With gest that the SPM‑2 shows a convergent relationship
the exception of Taste and Smell and Vision on the with a criterion measure of both the domain level of
Home Form, the SPM‑2 scales all showed significant sensory systems as well as the dimensional level of
correlations in the .50s between similar sensory sensory vulnerability.
scales. On the School Form, the correlations were
Validity SPM- 2 115

Table 5.16. Correlations Between SPM-2 and SP-2:
SPM-2 Preschool and SP-2 Child Forms
SP-2 Child Caregiver Forma

Aud Vis Tou Move Body Oral Cond Soc/Emo Attn
SPM-2 Preschool Home Form

Vision (VIS) .36 .36 .38 .46 .39 .34 .41 .42 .62
Hearing (HEA) .58 .07 .32 .35 .30 .34 .41 .43 .45
Touch (TOU) .44 .16 .54 .58 .21 .56 .59 .43 .47
Sensory Total (ST) .59 .26 .57 .62 .38 .28 .61 .54 .62
SP-2 Child Teacher Formb

Aud Vis Tou Move Beh – – – –
E
SPM-2 Preschool School Form
Vision (VIS) .53 .49 .60 .63 .46
PL .68 .68 .65
Touch (TOU) .47 .39 .67 .55 .54
Taste and Smell (T&S) .48 .22 .33 .22 .42
Body Awareness (BOD) .47 .51 .55 .62 .56
Balance and Motion (BAL) .51 .54 .51 .61 .50
Sensory Total (ST) .68 .59 .71 .72 .67
M
Planning and Ideas (PLN) .62 .58 .45 .56 .48
Social Participation (SOC) .60 .50 .48 .62 .73
Note. Corrected Pearson correlation coefficients, with significance at the p < .01 level indicated in bold. Aud = Auditory Processing; Vis = Visual
Processing; Tou = Touch Processing; Move = Movement Processing; Body = Body Position Processing; Oral = Oral Sensory Processing; Cond =
SA
Conduct associated with sensory processing; Soc/Emo = Social Emotional responses associated with sensory processing; Attn = Attentional
responses associated with sensory processing; Beh = Behavioral responses associated with sensory processing.
an = 85, ages 3 to 5 years. bn = 79, ages 3 to 5 years.

Table 5.17. Correlations Between SPM-2 and SP-2:
SPM-2 Child and SP-2 Child Forms
SP-2 Child Caregiver Forma

Aud Vis Tou Move Body Oral Cond Soc/Emo Attn
SPM-2 Child Home Form

Vision (VIS) .48 .55 .50 .41 .50 .32 .41 .51 .54
Hearing (HEA) .57 .42 .44 .37 .30 .32 .21 .38 .31
Touch (TOU) .34 .37 .54 .31 .41 .37 .33 .43 .34
Sensory Total (ST) .52 .48 .65 .55 .57 .54 .47 .52 .51
SP-2 Child Teacher Formb

Aud Vis Tou Move Beh – – – –
E
SPM-2 Child School Form
Vision (VIS) .47 .53 .43 .54 .45
Hearing (HEA) .55 PL .44 .55 .45 .48
Touch (TOU) .49 .50 .61 .55 .55
Taste and Smell (T&S) .27 .31 .39 .32 .39
Body Awareness (BOD) .48 .47 .59 .64 .58
Balance and Motion (BAL) .39 .39 .50 .54 .47
Sensory Total (ST) .56 .56 .64 .64 .61
M
Planning and Ideas (PLN) .55 .59 .49 .54 .59
Social Participation (SOC) .59 .53 .57 .54 .73
Note. Corrected Pearson correlation coefficients, with significance at the p < .01 level indicated in bold. Aud = Auditory Processing; Vis = Visual
Processing; Tou = Touch Processing; Move = Movement Processing; Body = Body Position Processing; Oral = Oral Sensory Processing; Cond =
SA
Conduct associated with sensory processing; Soc/Emo = Social Emotional responses associated with sensory processing; Attn = Attentional
responses associated with sensory processing; Beh = Behavioral responses associated with sensory processing.
Validity SPM- 2 117

Table 5.18. Correlations Between SPM-2 and SP:
SPM-2 Adolescent, SPM-2 Adult, and SP Adolescent/Adult Self-Report Forms
SP Adolescent/Adult Self-Report Form

Taste/Smell Move Vis Tou Activity Aud
SPM-2 Adolescent Self-Report Forma

Vision (VIS) .16 .12 .44 .23 .31 .29
Hearing (HEA) .26 .13 .19 .33 .33 .39
Touch (TOU) .25 .00 .09 .28 .27 .25
Taste and Smell (T&S) .43 .21 .04 .39 .27 .26
Body Awareness (BOD) .27 .26 −.01 .28 .16 .32
Balance and Motion (BAL) .33 .22 .20 .24 .19 .30
Sensory Total (ST) .36 .20 .20 .37 .31 .38
Planning and Ideas (PLN) .15 −.06 .26 .13 .16 .24
Social Participation (SOC) .23 .02 .27 .28 .29 .24
SPM-2 Adult Self-Report Formb

Vision (VIS) .23 .56 .63 .30 .21 .50
E
Hearing (HEA) .22 .55 .61 .45 .38 .80
Touch (TOU) .26 .38 .45 .51 .19 .37
Taste and Smell (T&S) .32
PL .44 .36 .43 .30 .26
Body Awareness (BOD) .29 .46 .45 .27 .26 .48
Balance and Motion (BAL) .37 .67 .59 .51 .26 .52
Sensory Total (ST) .35 .67 .68 .53 .35 .67
Planning and Ideas (PLN) .24 .34 .40 .18 .22 .25
Social Participation (SOC) .24 .41 .55 .32 .36 .43
M
Note. Corrected Pearson correlation coefficients, with significance at the p < .01 level indicated in bold. Taste/Smell = Taste/Smell Processing;
Move = Movement Processing; Vis = Visual Processing; Tou = Touch Processing; Activity = Activity Level; Aud = Auditory Processing.
SA

Convergence With the Adaptive Behavior Total (ST) correlation with the General Adaptive
Assessment System, Third Edition (ABAS‑3) Composite (GAC), the most general metrics on both
Sensory integration and processing is nomologically scales, ranged from −.43 to −.59, indicating a moder‑
related to adaptive behavior, or the “skills needed ate to strong negative relationship between sensory
to effectively and independently care for one‑ dysfunction and adaptive behavior. Furthermore,
self, respond to others, and meet environmental the GAC also showed moderate to strong significant
demands at home, school, work, and in the commu‑ correlations (−.29 to −.67) with the remaining SPM‑2
nity” (Harrison & Oakland, 2015, p. 1). In particular, scales, despite a substantial divergence in item con‑
adaptive behavior is compromised when a disability, tent across scales. Finally, particularly strong median
such as a sensory processing disorder, negatively correlations between scales with similar content
impacts the conceptual, social, and practical aspects indicated convergence between similar scales on
of daily living. Thus, it was expected that SPM‑2 both measures.
scores would correlate negatively with ABAS‑3 scores Overall, these results support the construct valid‑
because greater sensory integration and processing ity of the SPM‑2. The results also support the
dysfunction should be related to decreased adaptive criterion-related validity of the SPM‑2 in terms of its
behavior. capacity to detect deficits in adaptive functioning in
Tables 5.19 through 5.23 show the correlation clinical populations affected by sensory processing
matrices between the SPM‑2 and ABAS‑3 forms. As disorders.
E
expected, across all age levels, the median Sensory
Table 5.19. Correlations Between SPM-2 and ABAS-3:

PL
SPM-2 Infant and Toddler Forms
and ABAS-3 Parent/Primary Caregiver Form
ABAS-3 Parent/Primary Caregiver (Ages 0–5) Form
Com HS LS SC SD Soc MO GAC
SPM-2 Infant and Toddler Forms

M
Vision (VIS) −.43 −.33 −.45 −.34 −.33 −.42 −.34 −.43
Hearing (HEA) −.30 −.25 −.39 −.18 −.28 −.28 −.14 −.28
Touch (TOU) −.24 −.21 −.22 −.24 −.18 −.20 −.13 −.20
Taste and Smell (T&S) −.15 .05 −.02 −.14 −.08 −.02 .02 −.02
SA
Body Awareness (BOD) −.34 −.51 −.38 −.48 −.51 −.49 −.38 −.48
Balance and Motion (BAL) −.53 −.43 −.41 −.46 −.29 −.45 −.38 −.49
Sensory Total (ST) −.47 −.37 −.41 −.39 −.35 −.42 −.31 −.43
Planning and Ideas (PLN) −.58 −.53 −.52 −.66 −.45 −.59 −.60 −.67
Social Participation (SOC) −.37 −.50 −.52 −.58 −.51 −.58 −.42 −.53
Note. n = 52. Participant ages: 4 to 30 months. Corrected Pearson correlation coefficients, with significance at the p < .01 level indicated in bold.
Com = Communication; HS = Health and Safety; LS = Leisure; SC = Self-Care; SD = Self-Direction; Soc = Social; MO = Motor; GAC = General
Adaptive Composite.
Validity SPM- 2 119

SPM-2 Preschool Home and School Forms
and ABAS-3 Parent/Primary Caregiver and Teacher/Daycare Provider Forms
ABAS-3 Parent/Primary Caregiver and Teacher/Daycare Provider (Ages 2–5) Forms
Com FA SL/HLa HS LS SC SD Soc MO GAC
SPM-2 Preschool Home and

School Forms
Vision (VIS) −.25 −.24 −.28 −.44 −.34 −.24 −.37 −.37 −.34 −.40
Hearing (HEA) −.31 −.17 −.44 −.46 −.49 −.34 −.52 −.51 −.34 −.48
Touch (TOU) −.26 −.09 −.29 −.36 −.38 −.34 −.43 −.41 −.21 −.37
Taste and Smell (T&S) −.11 .00 −.11 −.13 −.19 −.15 −.20 −.14 −.08 −.15
Body Awareness (BOD) −.23 −.07 −.35 −.46 −.36 −.27 −.51 −.41 −.22 −.39
Balance and Motion (BAL) −.23 −.17 −.37 −.46 −.39 −.31 −.43 −.41 −.36 −.41
Sensory Total (ST) −.31 −.17 −.42 −.51 −.48 −.35 −.55 −.49 −.34 −.49
Planning and Ideas (PLN) −.55 −.42 −.52 −.65 −.63 −.48 −.60 −.63 −.49 −.67
Social Participation (SOC) −.34 −.19 −.52 −.55 −.60 −.40 −.74 −.60 −.27 −.58
Note. n = 181. Participant ages: 2 to 5 years. SPM-2 and ABAS-3 forms were each combined across home and school forms. Corrected Pearson
E
correlation coefficients, with significance at the p < .01 level indicated in bold. Com = Communication; FA = Functional Academics; SL/HL = School
Living / Home Living; HS = Health and Safety; LS = Leisure; SC = Self-Care; SD = Self-Direction; Soc = Social; MO = Motor; GAC = General Adaptive
Composite.
aOnly ABAS-3 scales present on both forms were included (except School Living and Home Living, which were combined).
PL
SPM-2 Child Home and School Forms
and ABAS-3 Parent and Teacher Forms
ABAS-3 Parent and Teacher (Ages 5–21) Forms
M
SPM-2 Child Home and
School Forms
Vision (VIS) −.45 −.42 −.37 −.33 −.39 −.39 −.41 −.42 −.46 −.45
Hearing (HEA) −.34 −.27 −.28 −.29 −.35 −.36 −.35 −.37 −.37 −.34
SA
Touch (TOU) −.40 −.31 −.35 −.37 −.42 −.47 −.43 −.42 −.45 −.40
Taste and Smell (T&S) −.29 −.14 −.20 −.27 −.32 −.37 −.25 −.31 −.30 −.29
Sensory Total (ST) −.49 −.38 −.40 −.42 −.49 −.51 −.46 −.49 −.52 −.49
Composite.

SPM-2 Adolescent Home and School Forms
and ABAS-3 Parent and Teacher Forms
ABAS-3 Parent and Teacher (Ages 5–21) Forms
SPM-2 Adolescent Home and

School Forms
Vision (VIS) −.55 −.49 −.44 −.49 −.47 −.40 −.51 −.57 −.57 −.55
Hearing (HEA) −.46 −.29 −.40 −.42 −.34 −.31 −.46 −.51 −.45 −.46
Touch (TOU) −.56 −.46 −.52 −.60 −.55 −.50 −.58 −.58 −.62 −.56
Sensory Total (ST) −.57 −.43 −.51 −.54 −.48 −.44 −.55 −.60 −.59 −.57
E
Composite.
PL
SPM-2 Adult Self-Report and Rater Report Forms
and ABAS-3 Adult Self-Report and Rated by Others Forms
ABAS-3 Adult Self-Report and Rated by Others (Ages 16–89) Forms
Com FA HL HS LS SC SD Soc MO GAC
M
SPM-2 Adult Self-Report and
Rater Report Forms
Vision (VIS) −.46 −.45 −.47 −.42 −.45 −.48 −.46 −.42 −.52 −.46
Hearing (HEA) −.43 −.26 −.33 −.24 −.33 −.31 −.31 −.28 −.37 −.43
SA
Touch (TOU) −.37 −.29 −.36 −.25 −.37 −.39 −.36 −.30 −.38 −.37
Sensory Total (ST) −.54 −.45 −.50 −.42 −.49 −.51 −.48 −.44 −.55 −.54
Note. n = 154. Participant ages: 21 to 67 years. Corrected Pearson correlation coefficients, with significance at the p < .01 level indicated in bold.
Com = Communication; FA = Functional Academics; HL = Home Living; HS = Health and Safety; LS = Leisure; SC = Self-Care; SD = Self-Direction;
Soc = Social; MO = Motor; GAC = General Adaptive Composite.
Validity SPM- 2 121

Validity Evidence Based on Clinical Groups Effect Sizes in the General Clinical Sample
Criterion-related validity refers to how well a test pre‑ A series of matched-sample studies (total N = 580)
dicts an independent outcome, or criterion. For the across all five age levels examined the extent to
SPM‑2, an important index of criterion validity is the which typical and clinical group participants showed
capacity to distinguish between groups of children meaningful differences on SPM‑2 variables. Par‑
who are expected to differ in sensory processing ticipants from the clinical group were matched to a
function, praxis, and social participation. Unlike the randomly selected subsample from the standardiza‑
large, demographically representative SPM‑2 stan‑ tion sample on the basis of age, gender, education
dardization samples, the clinical groups are smaller, level, and ethnicity. The matched sampling was
less broadly representative samples collected by accomplished by using the propensity score approach
clinicians in a variety of settings. The demographic (nearest neighbor method), as implemented in the
characteristics of these groups are summarized in R package “MatchIt” (Ho et al., 2007, 2011).
Table 4.13.
Tables 5.24 through 5.28 show the means, standard
In the sections that follow, SPM‑2 score comparisons deviations, and effect sizes for the SPM‑2 scales in
are presented in tables that display the descriptive the clinical and matched typical (standardization)
statistics for the groups being compared, along with samples. On all forms, the scale means are higher
the effect size (Cohen’s d) of the difference between (indicating more symptoms) in the clinical sample
E
the mean scores of each group. By convention, effect than in the typical sample. The Sensory Total (ST)
sizes of d ≥ 0.40 are deemed to represent clinically scores, for example, show strong effect sizes in the
significant effects, meaning that the score differ‑ .40 to .50 range for the Infant/Toddler and Preschool
ences signify behavioral differences that are likely PL age levels, and in the .50 to .80 range for the Child
to be observable in individuals, not just in aggregate and Adolescent age levels. At the Adult age level, the
results from a larger group. effect was more modest, in the .20 to .30 range. These
clinically meaningful effect sizes demonstrate that
the SPM‑2 scales can reliably distinguish between
typically developing and clinic-referred clients.
M
Table 5.24. Mean SPM-2 T-Scores and Effect Sizes, Mixed Clinical Sample and Matched Control Group:
Infant/Toddler Age Level Forms
Mixed clinical sample Matched control group

Scale Mean SD Mean SD Effect size
SA
Infant and Toddler Forms

Vision (VIS) 51.1 11.1 47.5 10.7 0.22
Hearing (HEA) 52.0 10.4 47.6 9.7 0.30
Touch (TOU) 51.7 10.0 48.4 9.4 0.23
Taste and Smell (T&S) 50.8 11.7 48.2 9.4 0.17
Body Awareness (BOD) 53.4 11.2 48.2 9.9 0.34
Balance and Motion (BAL) 59.0 12.4 48.2 10.6 0.64
Sensory Total (ST) 53.7 10.7 47.4 10.3 0.40
Planning and Ideas (PLN) 63.0 13.6 49.3 9.6 0.84
Social Participation (SOC) 56.3 11.8 48.9 9.7 0.48
Note. Combined Infant and Toddler Forms n = 54. Caregiver Self-Report Forms were not administered to a clinical sample. Effect size (Cohen’s d) = the
scale mean in the clinical sample minus the scale mean in the control group sample, divided by the pooled standard deviation.

Preschool Age Level Forms

Home Form
Vision (VIS) 53.9 10.3 49.5 9.0 0.32
Hearing (HEA) 55.8 9.7 49.9 9.2 0.42
Touch (TOU) 55.9 9.8 50.5 9.5 0.38
Sensory Total (ST) 56.3 9.7 50.6 9.2 0.40
School Form
Vision (VIS) 56.5 10.7 49.4 10.2 0.46
E
Hearing (HEA) 57.4 9.7 49.3 9.5 0.56
Touch (TOU) 55.7 9.4 50.6 9.3 0.37
Sensory Total (ST)
PL
55.8
55.9
57.4
10.1
10.8
9.9
51.2
50.1
50.0
9.1
9.9
9.9
0.33
0.38
0.49
M
Note. Home Form n = 107; School Form n = 99. Effect size (Cohen’s d) = the scale mean in the clinical sample minus the scale mean in the control
group sample, divided by the pooled standard deviation.
SA
Validity SPM- 2 123

Child Age Level Forms

Home Form
Vision (VIS) 60.0 9.9 50.3 9.1 0.70
Hearing (HEA) 58.9 9.8 50.5 8.9 0.61
Touch (TOU) 59.8 10.1 50.8 9.3 0.62
Sensory Total (ST) 61.4 9.5 50.3 9.7 0.77
School Form
Vision (VIS) 62.9 11.2 49.7 9.9 0.86
E
Hearing (HEA) 65.5 10.2 50.2 9.7 1.03
Touch (TOU) 62.5 11.6 50.0 9.1 0.84
Sensory Total (ST)
58.2
62.2
63.8
PL 12.2
11.7
10.8
50.4
50.1
50.1
9.3
9.4
9.3
0.51
0.79
0.93
M
Note. Home Form n = 218; School Form n = 194. Effect size (Cohen’s d) = the scale mean in the clinical sample minus the scale mean in the control
group sample, divided by the pooled standard deviation.
SA

Adolescent Age Level Forms

Home Form
Vision (VIS) 59.0 8.1 48.3 7.7 0.92
Hearing (HEA) 58.9 8.9 49.1 8.5 0.75
Touch (TOU) 57.8 9.3 49.1 7.9 0.69
Sensory Total (ST) 58.9 8.0 48.5 8.7 0.83
School Form
Vision (VIS) 56.4 8.7 50.3 7.8 0.51
E
Hearing (HEA) 57.9 8.2 49.6 7.6 0.71
Touch (TOU) 57.8 8.8 49.9 7.0 0.69
Sensory Total (ST)
PL
57.4
56.7
58.8
8.6
8.7
8.0
50.2
49.7
49.3
7.3
6.5
8.8
0.62
0.64
0.74
M
Self-Report Form
Vision (VIS) 56.8 10.3 47.6 8.5 0.68
Hearing (HEA) 56.6 9.1 46.9 8.3 0.75
Touch (TOU) 55.0 10.3 47.7 7.9 0.56
SA

Sensory Total (ST) 56.4 9.9 46.7 8.2 0.73
Note. Home Form n = 105; School Form n = 102; Self-Report Form n = 105. Effect size (Cohen’s d) = the scale mean in the clinical sample minus the
scale mean in the control group sample, divided by the pooled standard deviation.
Validity SPM- 2 125

Adult Age Level Forms

Self-Report Form
Vision (VIS) 55.9 11.4 50.5 9.7 0.35
Hearing (HEA) 54.7 11.9 50.1 9.9 0.29
Touch (TOU) 53.6 10.2 49.6 10.8 0.26
Sensory Total (ST) 54.2 11.2 50.3 10.4 0.25
Rater Report Form

Vision (VIS) 55.9 10.9 51.5 9.0 0.30
E
Hearing (HEA) 54.2 10.8 50.8 10.2 0.21
Touch (TOU) 52.0 9.5 51.1 9.5 0.07
Sensory Total (ST)
53.1
53.6
54.3
PL 10.3
10.8
10.3
51.0
50.7
51.1
8.7
9.7
10.2
0.16
0.19
0.21
M
Note. Self-Report Form n = 116; Rater Report Form n = 120. Effect size (Cohen’s d) = the scale mean in the clinical sample minus the scale mean in the
control group sample, divided by the pooled standard deviation.
SA

Effect Sizes by Clinical Disorder compared to the typically developing individuals
in the SPM‑2 standardization sample. The sensory
Of greater interest are specific clinical disorders
processing and autism groups, in particular, show a
within the general clinic-referred sample. Tables 5.29
pattern of large effects, indicating that the SPM‑2 is
to 5.34 show SPM‑2 scale descriptive statistics and
particularly sensitive to sensory integration and pro‑
effect sizes for six diagnostic groups with adequate
cessing dysfunction. The intellectual disability group
sample sizes to permit additional analysis: sensory
also shows moderate to large effects, with expected
processing disorders (SPD), autism spectrum disor‑
peaks in planning and ideation. The speech pathol‑
ders (ASD), intellectual and developmental disorders
ogy group showed small to medium effects, with
(ID/DD), speech–language pathologies (SLP),
larger mean score differences in vision and planning
attention-deficit/hyperactivity disorder (ADHD),
and ideation. The mental health disorders groups,
and mood and anxiety disorders (MAD). All of these
composed of participants with ADHD and mood
disorders frequently co-occur with various deficits in
and anxiety disorders, showed clinically meaning‑
sensory integration and processing function.
ful effect sizes across SPM-2 scales in the small to
As a whole, the diagnostic groups show clinically medium range.
meaningful elevations in SPM‑2 scale scores when
Table 5.29. Mean SPM-2 T-Scores and Effect Sizes
E
for Sensory Processing Disorders (SPD) and Matched Control Groups
SPD group Matched control group

Home Forms
Vision (VIS)
Hearing (HEA)
PL
60.4
60.4
9.9
10.3
47.9
50.1
8.2
8.2
0.96
0.77
Touch (TOU) 60.9 10.0 48.4 8.5 0.92
M
Sensory Total (ST) 63.1 8.4 49.4 7.2 1.20
SA
School Forms
Vision (VIS) 56.3 10.2 49.2 7.6 0.56
Hearing (HEA) 59.7 10.1 50.4 9.1 0.66
Touch (TOU) 59.5 9.5 50.4 9.7 0.63
Sensory Total (ST) 59.8 9.6 49.8 8.6 0.75
Note. Home Forms (n = 39) include clinical participants from the Infant/Toddler, Preschool, Child, and Adolescent age levels. School Forms (n = 34)
include clinical participants from the Preschool, Child, and Adolescent age levels. Matched control group participants were randomly drawn from the
corresponding standardization samples based on matching demographic variables. Effect size (Cohen’s d) = the scale mean in the clinical sample
minus the scale mean in the standardization sample, divided by the pooled standard deviation.
Validity SPM- 2 127

for Autism Spectrum Disorders (ASD) and Matched Control Groups
ASD group Matched control group

Home Forms
Vision (VIS) 62.5 8.4 49.0 9.6 0.97
Hearing (HEA) 63.2 6.5 48.9 9.3 1.14
Touch (TOU) 62.7 7.3 50.4 9.1 0.97
Sensory Total (ST) 64.9 6.6 49.1 9.6 1.21
School Forms
Vision (VIS) 64.2 8.3 47.9 8.0 1.35
E
Hearing (HEA) 65.5 8.5 48.2 7.3 1.50
Touch (TOU) 64.7 9.0 48.8 7.4 1.34
Sensory Total (ST)
63.6
61.6
66.0
PL 9.3
10.3
7.9
49.1
49.7
48.3
8.1
8.4
7.8
1.13
0.88
1.52
M
Self-Report Forms
Vision (VIS) 58.3 9.2 50.0 6.5 0.75
Hearing (HEA) 57.9 9.2 49.5 7.9 0.67
Touch (TOU) 57.4 7.5 48.2 7.2 0.84
SA

Sensory Total (ST) 58.0 8.1 48.1 7.0 0.90
include clinical participants from the Preschool, Child, and Adolescent age levels. Self-Report Forms (n = 35) include clinical participants from the
Adolescent and Adult age levels. Matched control group participants were randomly drawn from the corresponding standardization samples based on
matching demographic variables. Effect size (Cohen’s d) = the scale mean in the clinical sample minus the scale mean in the standardization sample,
divided by the pooled standard deviation.

for Intellectual and Developmental Disabilities (ID/DD) and Matched Control Groups
ID/DD group Matched control group

Home Forms
Vision (VIS) 56.1 9.2 47.6 10.2 0.57
Hearing (HEA) 54.8 9.4 48.0 8.9 0.49
Touch (TOU) 58.5 10.4 47.9 8.8 0.76
Sensory Total (ST) 58.1 8.8 48.4 9.3 0.71
School Forms
Vision (VIS) 62.0 10.0 48.7 8.0 1.02
E
Hearing (HEA) 59.0 9.5 47.7 7.3 0.93
Touch (TOU) 60.7 9.8 49.1 7.0 0.98
Sensory Total (ST)
PL
59.3
57.8
61.1
9.5
10.5
8.7
50.0
48.5
48.7
7.6
7.6
7.0
0.75
0.72
1.09
M
include clinical participants from the Preschool, Child, and Adolescent age levels. Matched control group participants were randomly drawn from the
corresponding standardization samples based on matching demographic variables. Effect size (Cohen’s d) = the scale mean in the clinical sample
minus the scale mean in the standardization sample, divided by the pooled standard deviation.
SA
Validity SPM- 2 129

for Speech–Language Pathologies (SLP) and Matched Control Groups
SLP group Matched control group

Home Forms
Vision (VIS) 52.5 10.4 48.2 9.4 0.29
Hearing (HEA) 51.0 10.0 48.1 9.6 0.20
Touch (TOU) 52.9 9.0 48.8 8.9 0.31
Sensory Total (ST) 53.3 9.8 48.0 9.3 0.38
School Forms
Vision (VIS) 57.9 12.1 49.0 8.8 0.60
E
Hearing (HEA) 56.1 10.1 49.1 8.0 0.54
Touch (TOU) 56.9 10.7 48.9 8.0 0.60
Sensory Total (ST)
56.8
57.4
57.7
PL 11.9
11.0
11.7
49.0
49.9
48.4
8.3
7.9
8.8
0.55
0.57
0.63
M
Self-Report Forms
Vision (VIS) 56.4 10.9 45.3 8.8 0.78
Hearing (HEA) 52.9 9.5 47.7 10.4 0.34
Touch (TOU) 53.0 9.3 49.6 9.2 0.24
SA

Sensory Total (ST) 54.4 8.5 46.9 9.2 0.56
include clinical participants from the Preschool, Child, and Adolescent age levels. Self-Report Forms (n = 14) include clinical participants from the
Adolescent and Adult age levels. Matched control group participants were randomly drawn from the corresponding standardization samples based on
matching demographic variables. Effect size (Cohen’s d) = the scale mean in the clinical sample minus the scale mean in the standardization sample,
divided by the pooled standard deviation.

for Attention-Deficit/Hyperactivity Disorder (ADHD) and Matched Control Groups
ADHD group Matched control group

Home Forms
Vision (VIS) 59.3 7.7 50.2 7.9 0.77
Hearing (HEA) 58.6 8.1 50.8 8.5 0.62
Touch (TOU) 57.8 9.5 50.4 8.4 0.57
Sensory Total (ST) 59.9 7.1 50.3 8.7 0.79
School Forms
Vision (VIS) 58.4 9.6 49.2 7.2 0.77
E
Hearing (HEA) 60.1 9.9 48.6 7.9 0.89
Touch (TOU) 58.6 9.6 48.1 7.8 0.83
Sensory Total (ST)
PL
59.8
57.7
60.2
10.5
9.6
9.8
49.7
50.0
49.0
8.7
8.6
8.9
0.72
0.58
0.82
M
Note. Home Forms (n = 57) include clinical participants from the Preschool, Child, and Adolescent age levels. School Forms (n = 44) include clinical
participants from the Preschool, Child, and Adolescent age levels. Matched control group participants were randomly drawn from the corresponding
standardization samples based on matching demographic variables. Effect size (Cohen’s d) = the scale mean in the clinical sample minus the scale
mean in the standardization sample, divided by the pooled standard deviation.
SA
Validity SPM- 2 131

for Mood and Anxiety Disorders (MAD) and Matched Control Groups
MAD group Matched control group

Home Forms
Vision (VIS) 57.4 8.8 48.3 9.9 0.63
Hearing (HEA) 57.5 8.7 50.1 9.4 0.54
Touch (TOU) 56.1 8.2 48.7 9.9 0.53
Sensory Total (ST) 57.4 7.9 47.8 11.2 0.64
School Forms
Vision (VIS) 55.8 10.3 49.2 8.2 0.50
E
Hearing (HEA) 57.7 9.5 49.4 8.9 0.61
Touch (TOU) 58.2 10.5 48.9 6.2 0.81
Sensory Total (ST)
56.0
54.3
57.9
PL 11.0
10.3
10.0
49.4
48.9
49.0
7.7
7.0
8.1
0.50
0.45
0.68
M
Note. Home Forms (n = 42) include clinical participants from the Preschool, Child, and Adolescent age levels. School Forms (n = 48) include clinical
participants from the Preschool, Child, and Adolescent age levels. Typical participants were randomly drawn from the corresponding standardization
samples based on matching demographic variables. Mood and anxiety disorders were 50% mood disorders and 50% anxiety disorders. Effect size
(Cohen’s d) = the scale mean in the clinical sample minus the scale mean in the standardization sample, divided by the pooled standard deviation.
SA

Detection of Sensory Processing Disorders question. Specificity, on the other hand, refers to a
scale’s ability to exclude cases that do not have the
Disorders of sensory integration and processing rep‑
disorder in question. To illustrate, Table 5.35 shows
resent the clinical conditions that are most directly
that at a cutoff score of T = 60, the ST scale has sensi‑
targeted by the SPM‑2 sensory systems scales. The
tivity of .72 and specificity of .83. In practical terms,
SPM‑2 clinical sample included a subsample (n = 39)
this means that 72% of the individuals with sensory
of individuals whose primary diagnosis or clinical
processing disorders had ST scores of 60 or greater,
condition was identified as a disorder of sensory
whereas 83% of the typically developing individuals
integration or sensory processing. Conditional prob‑
had scores of 59 or less.
ability analyses (area under the curve, AUC) were
run to determine the ability of the SPM‑2 Sensory Table 5.35 serves as a reminder that at any level of
Total (ST) scale to detect these conditions at vari‑ score elevation, there is a risk of under- or over‑
ous cutoff points. As previously noted, the ST scale identifying cases of sensory processing disorder. In
is a 60‑item composite of all of the SPM‑2 sensory support of the criterion-related validity of the SPM‑2
systems items. The ST score provided statistically ST scale, it is worth noting that only 2% or less of
significant improvement over chance in detecting typically developing individuals had T‑scores of 70
these disorders. or higher on any SPM-2 form (specificity = .98). This
finding demonstrates that when an individual’s
Table 5.35 displays the sensitivity and specificity
score is in the Severe Difficulties interpretive range
E
associated with various values of the ST T‑score, for
(T ≥ 70), there is a strong probability that the
all SPM‑2 forms. Sensitivity refers to a rating scale’s
individual has clinically significant problems with
ability to detect true positive cases of the disorder in
sensory processing or integration.
cutoff points
PL
Table 5.35. Conditional Probability Analysis for Detection of Sensory Processing Disorders
Sensory Total (ST) T-score
Sensitivity Specificity
55 .87 .67
60 .72 .83
M
65 .51 .92
70 .23 .98
75 .03 .99
Note. Sample included 39 individuals with sensory processing/integration disorders and 3,096 typically developing
SA
individuals from the standardization sample; analysis included every form.
Validity SPM- 2 133

Detection of Differential Functioning Between In these distributions, the No Difference range (–9
Environments and Raters to 9, or within 1 standard deviation) represents dif‑
ferences between the two scores that are within the
The Difference (DIF) score was developed to quantify
normal range of variability. The Probable Difference
score differences between the same scales (e.g., VIS,
range (–10 to –14 and 10 to 14, or between 1 and
HEA, TOU) on different forms. The DIF score is cal‑
1.5 standard deviations) and the Definite Difference
culated by subtracting the T‑score of one form from
range (≤ –15 and ≥ 15, or equal to or greater than 1.5
the comparable T‑score of another, for example, the
standard deviations) represent two levels of certainty
Child School ST T‑score from the Child Home ST
about the difference between the scores.
T‑score. (See Chapter 2 for a detailed description of
scoring procedures.) Because each T‑score is normal‑ These interpretive ranges were set to approximate
ized, the DIF score distribution is also approximately conventional cutoffs for interpreting T‑scores as
normal. Table 5.36 shows the distributions of the ST typical, mildly atypical, and moderately to extremely
scores, in both the standardization and clinical sam‑ atypical. The sign of the DIF score is relevant only
ples, for individuals assessed in the home and school for the Probable Difference and Definite Difference
environments (Preschool, Child, Adolescent age ranges, with a positive or negative score indicating
levels); for individuals assessed by self-report and more problems reported on one form or the other.
rated by others (Adolescent, Adult age levels); and
for individuals assessed in conjunction with another
E
individual (Infant/Toddler age level, Caregiver).
Table 5.36. Frequencies of Sensory Total (ST) Difference (DIF) Scores
DIF score interpretive ranges

PL
in the SPM-2 Standardization and Clinical Samples
Standardization sample
n % of sample n
Clinical sample
% of sample
Infant/Toddler age level

Infant Form vs. Caregiver Self-Report Form
≤ –15 Definite Difference 10 3.9 0 0.0
M
–14 to –10 Probable Difference 21 8.3 2 18.2
–9 to 9 No Difference 182 71.7 6 54.5
10 to 14 Probable Difference 23 9.1 2 18.2
SA
≥ 15 Definite Difference 18 7.1 1 9.1
Toddler Form vs. Caregiver Self-Report Form

–9 to 9 No Difference 297 73.7 25 58.1
Preschool age level

Home Form vs. School Form
–9 to 9 No Difference 201 67.0 60 61.2
Note. Sample size discrepancies are due to missing data.

Table 5.36. Frequencies of Sensory Total (ST) Difference (DIF) Scores
in the SPM-2 Standardization and Clinical Samples (continued)
DIF score interpretive ranges n % of sample n % of sample
Child age level

–9 to 9 No Difference 271 65.3 112 61.2
Adolescent age level

–9 to 9 No Difference 398 67.0 62 74.7
E
Home Form vs. Self-Report Form

≤ –15
–14 to –10
–9 to 9
10 to 14
Definite Difference
Probable Difference
No Difference
Probable Difference
PL 37
55
538
69
5.1
7.6
74.3
9.5
2
3
59
13
70.2
15.5
2.4
3.6
School Form vs. Self-Report Form

M
–9 to 9 No Difference 325 55.7 44 52.4
SA

Adult age level

Self-Report Form vs. Rater Report Form
–9 to 9 No Difference 314 66.2 88 78.6
Note. Sample size discrepancies are due to missing data.
Validity SPM- 2 135

Summary
This chapter has reported on the initial psychometric across the SPM‑2 forms. The SPM‑2 forms correlate
properties of the SPM‑2. The SPM‑2 forms perform in expected ways with two measures of sensory inte‑
well on seven key indices of reliability: internal gration and processing, the Sensory Profile and the
consistency, test–retest reliability, standard error of Sensory Profile 2, and a measure of adaptive behav‑
measurement, interrater reliability, alternate-forms ior, the ABAS‑3. Finally, the SPM‑2 forms distinguish
reliability, cross-rater concordance, and online and between typically developing and clinic-referred
paper forms equivalence. Analyses of content and clients, with robust and clinically meaningful effect
structure support the scoring of separate sensory sizes. The authors of the SPM‑2 hope that its publica‑
systems, praxis, and social participation scores tion will stimulate additional research.
E
PL
M
SA

05 spm2 Manual CH Sample 041621 1

Uploaded by

Copyright:

Available Formats

05 spm2 Manual CH Sample 041621 1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

05 spm2 Manual CH Sample 041621 1

Uploaded by

Copyright:

Available Formats

Selected material from the Sensory Processing Measure, Second Edition (SPMTM-2) manual.

Copyright © 2021 by Western Psychological Services (WPS®).

SPM-2 • W-706M wpspublish.com

05_SPM_2_Manual_CH5_032421.indd 95 4/1/21 12:23 PM

bility, refers to the stability of test scores over time.

96 SPM- 2 Chapter 5 Psychometric Properties

SPM-2 • W-706M wpspublish.com

05_SPM_2_Manual_CH5_032421.indd 96 4/1/21 12:23 PM

Internal consistencya Test–retest reliabilityb

Age group (months)

Caregiver Self-Report Form

Note. CI = confidence interval.

SPM-2 • W-706M wpspublish.com

05_SPM_2_Manual_CH5_032421.indd 97 4/1/21 12:23 PM

Internal consistencya Test–retest reliabilityb

Note. CI = confidence interval.

bTwo-week test–retest correlation; Home Form n = 77; School Form n = 92.

98 SPM- 2 Chapter 5 Psychometric Properties

SPM-2 • W-706M wpspublish.com

05_SPM_2_Manual_CH5_032421.indd 98 4/1/21 12:23 PM

Internal consistencya Test–retest reliabilityb

Note. CI = confidence interval.

bTwo-week test–retest correlation; Home Form n = 129; School Form n = 110.

SPM-2 • W-706M wpspublish.com

05_SPM_2_Manual_CH5_032421.indd 99 4/1/21 12:23 PM

Internal consistencya Test–retest reliabilityb

Note. CI = confidence interval.

100 SPM- 2 Chapter 5 Psychometric Properties

SPM-2 • W-706M wpspublish.com

05_SPM_2_Manual_CH5_032421.indd 100 4/1/21 12:23 PM

Rater Report Form

Child School Environment Forms

Adolescent and Adult Driving Environment Forms

102 SPM- 2 Chapter 5 Psychometric Properties

SPM-2 • W-706M wpspublish.com

05_SPM_2_Manual_CH5_032421.indd 102 4/1/21 12:23 PM

These results are consistent with other behavior

Reliability SPM- 2 103

SPM-2 • W-706M wpspublish.com

05_SPM_2_Manual_CH5_032421.indd 103 4/1/21 12:23 PM

104 SPM- 2 Chapter 5 Psychometric Properties

SPM-2 • W-706M wpspublish.com

05_SPM_2_Manual_CH5_032421.indd 104 4/1/21 12:23 PM

psychometric characteristics. internal consistency (alpha) coefficients in Tables 5.1

Validity SPM- 2 105

SPM-2 • W-706M wpspublish.com

05_SPM_2_Manual_CH5_032421.indd 105 4/1/21 12:23 PM

Touch (TOU) .57 .63 — — — — — — —

106 SPM- 2 Chapter 5 Psychometric Properties

SPM-2 • W-706M wpspublish.com

05_SPM_2_Manual_CH5_032421.indd 106 4/1/21 12:23 PM

Validity SPM- 2 107

SPM-2 • W-706M wpspublish.com

05_SPM_2_Manual_CH5_032421.indd 107 4/1/21 12:23 PM

108 SPM- 2 Chapter 5 Psychometric Properties

SPM-2 • W-706M wpspublish.com

05_SPM_2_Manual_CH5_032421.indd 108 4/1/21 12:23 PM

Touch (TOU) .70 .74 — — — — — — —

Validity SPM- 2 109