Psychological Reports, 2012, 110, 1, 197-217.

© Psychological Reports 2012

THE CHILDREN’S BEHAVIOR QUESTIONNAIRE VERY SHORT

SCALE: PSYCHOMETRIC PROPERTIES AND DEVELOPMENT
OF A ONE-ITEM TEMPERAMENT SCALE1, 2

ESTER F. C. SLEDDENS
Department of Health Promotion
NUTRIM School for Nutrition, Toxicology and Metabolism
Maastricht University Medical Center+

SHERYL O. HUGHES, TERESIA M. O’CONNOR, ALICIA BELTRAN, JANICE C.

BARANOWSKI, THERESA A. NICKLAS, AND TOM BARANOWSKI
Department of Pediatrics
Baylor College of Medicine
USDA/ARS Children’s Nutrition Research Center

Summary.—Little research has been conducted on the psychometrics of the very

short scale (36 items) of the Children’s Behavior Questionnaire, and no one-item
temperament scale has been tested for use in applied work. In this study, 237 United
States caregivers completed a survey to define their child’s behavioral patterns (i.e.,
Surgency, Negative Affectivity, Effortful Control) using both scales. Psychometrics
of the 36-item Children’s Behavior Questionnaire were examined using classical test
theory, principal factor analysis, and item response modeling. Classical test theory
analysis demonstrated adequate internal consistency and factor analysis confirmed
a three-factor structure. Potential improvements to the measure were identified us-
ing item response modeling. A one-item (three response categories) temperament
scale was validated against the three temperament factors of the 36-item scale. The
temperament response categories correlated with the temperament factors of the
36-item scale, as expected. The one-item temperament scale may be applicable for
clinical use.

Children make important contributions to their interpersonal relation-

ships. Child temperament is one child characteristic that has been shown
to strongly affect the development of appropriate social interactions in
children (Putnam, Sanson, & Rothbart, 2002). Temperament has been de-
fined as a function of biological make-up, reactivity to stimuli (arousabil-
1
Address correspondence to Ester F. C. Sleddens, Department of Health Promotion, NU-
TRIM School for Nutrition, Toxicology and Metabolism, Maastricht University Medical
Center+, P.O. Box 616, 6200 MD, Maastricht, The Netherlands, or e-mail (ester.sleddens@
maastrichtuniversity.nl).
2
This research was primarily funded by a grant from the National Institute of Child Health
and Human Development (HD058175) for use by the United States Department of Agricul-
ture (USDA/ARS) Children’s Nutrition Research Center, Department of Pediatrics, Baylor
College of Medicine, Houston, Texas, and was funded in part with federal funds from the
USDA/ARS under Cooperative Agreement No. 58-6250-6001. The contents of this publica-
tion do not necessarily reflect the views or policies of the USDA, nor does mention of trade
names, commercial products, or organizations imply endorsement from the United States
government. This research was also partially funded by the Netherlands Heart Foundation
(Project Number 2008B112) to Ester Sleddens.

DOI 10.2466/08.10.21.PR0.110.1.197-217 ISSN 0033-2941

198 E. F. C. SLEdDens, et al.

ity of motor, affective, and sensory response systems), and self-regulation

(including attentional strategies and behavioral avoidance) (Rothbart &
Bates, 1998; Rothbart, Ahadi, Hershey, & Fisher, 2001). The Children’s Be-
havior Questionnaire (Rothbart, et al., 2001) is one of the most well-known
and comprehensive instruments assessing individual differences in facets
of temperament in children ages three to eight years old.
Progressively shorter versions of the Children’s Behavior Question-
naire have been validated for use in research, including the standard (195
items), short (94 items), and very short scales (36 items) (Rothbart, et al.,
2001; Putnam & Rothbart, 2006). All scales capture three broad tempera-
mental traits: Surgency or Extraversion (in this paper referred to as Sur-
gency), Negative Affectivity, and Effortful Control, which are conceptual-
ly similar to three of the major “Big Five” dimensions of adult personality,
labeled Extraversion, Neuroticism, and Conscientiousness, respectively
(Rothbart, Ahadi, & Evans, 2000). Whereas the first two traits are charac-
terized by behaviors reactive to environmental changes, the third is the
self-regulation aspect of temperament, involving processes that moderate
reactivity. These higher order temperament factors have been repeatedly
confirmed using factor analytic procedures in the United States (Rothbart,
Putnam & Rothbart, 2006) and European (Nygaard, Smith, & Torgersen,
2002; Van den Bergh & Ackx, 2003, Majdandžić & Van den Boom, 2007;
Eggers, De Nil, & Van den Bergh, 2009, 2010; Sleddens, Kremers, Candel,
De Vries, & Thijs, 2011) and Asian countries (Ahadi, Rothbart, & Ye, 1993;
Rothbart, et al., 2001).
Although the standard scale of the Children’s Behavior Questionnaire
gives better estimates of temperamental behavior patterns, administering
the longer scale is not always feasible due to time constraints or partici-
pant burden. The three higher-order temperament factors, however, have
been invariant across the two shorter versions (Putnam & Rothbart, 2006;
Sleddens, et al., 2011), and other psychometric results (e.g., internal relia-
bility coefficients) were promising for the 94-item scale (Putnam & Roth-
bart, 2006; Healey, Brodzinsky, Bernstein, Rabinovitz, & Halperin, 2010;
Tumanova, Zebrowski, Throneburg, & Kulak Kayikci, 2011) and the 36-
item scale (Putnam & Rothbart, 2006; Hughes, Shewchuk, Baskin, Nick-
las, & Qu, 2008; Ward, Gay, Alkon, Anders, & Lee, 2008; Miller, Vannatta,
Compas, Vasey, McGoron, Salley, et al., 2009; O’Connor, et al., 2010; Sled-
dens, et al., 2011).
The very short scale of the Children’s Behavior Questionnaire, con-
sisting of items extracted from the standard scale, has been more frequent-
ly administered during the last few years (Hughes, et al., 2008; Ward, et al.,
2008; Miller, et al., 2009; O’Connor, et al., 2010). However, the factorial va-
lidity of this 36-item scale has only been established when administered as
part of the longer scales of the Children’s Behavior Questionnaire (Putnam
 Validation of CBQ and One-Item Temperament Scale 199

& Rothbart, 2006; Sleddens, et al., 2011). In addition, advanced psychomet-

ric procedures, e.g., item response modeling, have not been reported for
the Children’s Behavior Questionnaire. In the current study, the 36-item
scale was administered to evaluate its psychometric properties using clas-
sical test theory (internal consistency and corrected item-total correlation)
and factor analysis (Putnam & Rothbart, 2006) in an ethnically diverse
sample. Item response modeling was also applied to extend the classical
test theory approach, a method which examines whether the item content
covers the range of respondents’ answers about the construct of interest
(Heesch, Mâsse, & Dunn, 2006; Wilson, Allen, & Li, 2006a, 2006b).
It has been shown that an incompatibility or “poor fit” between the
temperamental characteristics of the child and expectations and attitudes
of caregivers may lead to poor outcomes for children (Putnam, et al., 2002).
Interventions targeting parenting may benefit from being tailored to the
child’s temperament (e.g., Kochanska, Aksan, & Joy, 2007; Van Zeijl, Mes-
man, Stolk, Alink, Van Ijzendoorn, Bakermans-Kranenburg, et al., 2007;
Bradley & Corwyn, 2008; Zeller, Boles, & Reiter-Purtill, 2008; Gubbels, Kre-
mers, Stafleu, Dagnelie, Goldbohm, De Vries, et al., 2009; Wu, Dixon, Dal-
ton, Tudiver, & Liu, 2010), but a simple way to assess a child’s tempera-
ment would be needed. For example, researchers using game technology
to develop an interactive electronic game to teach parents strategies to in-
crease vegetable consumption in preschoolers used a short assessment of
child temperament to tailor parenting strategies to the temperament of
their child (Baranowski, O’Connor, Hughes, Beltran, Baranowski, Nicklas,
et al., in press; Beltran, O’Connor, Hughes, Baranowski, Nicklas, Thomp-
son, et al., in press). Such a game that simulates a child with whom parents
can interact during feeding scenarios may benefit from having the child
character reflect the temperament of the player’s child, making the interac-
tion more realistic. Within the context of a video game, the assessment of
the player’s child’s temperament needs to be done rapidly, or else the play-
er’s attention and interest will be lost, and parents need to identify the sin-
gle temperament characteristic they feel is most dominant in their child to
make tailoring of the game feasible. A single-item assessment of the play-
er’s child’s temperament would work well in such a situation, but the sin-
gle item must be validated. This approach increases the likelihood of par-
ents participating in the game and benefiting from the strategies proposed.
Hence, besides assessing the psychometrics of the 36-item temperament
scales, this study aimed to assess the validity of a one-item scale, where-
in parents chose from one of three vignettes to describe their child’s tem-
perament. It was hypothesized that parents selecting the one best vignette
to describe the dominant temperament characteristic of their child would
provide the highest mean scores on the corresponding temperament factor
when rating their child on the 36-item scale.
200 E. F. C. SLEdDens, et al.

Method
Respondents and Procedure
United States English-speaking caregivers of children ages 3 to 5 years
were recruited for this cross-sectional online survey about child tempera-
ment, in which both the 36-item and the one-item temperament scale were
administered. Potential participants were approached from July 2010 un-
til February 2011 by (a) posting flyers in the vicinity of the Texas Medical
Center, community centers, public libraries, universities, sports centers,
and museums throughout Houston, Texas; (b) posting the study on the
website of Baylor College of Medicine and the Children’s Nutrition Re-
search Center (CNRC); and (c) listing the study in the CNRC’s nationally
distributed newsletter and recruiting from the participant database. Com-
pletion time for the survey was approximately 10 minutes. From all com-
pleted entries, two drawings were conducted for two $100 gift cards. Only
participants who agreed to take part in the raffles (n = 224) had a chance to
win one of the gift cards.
A total of 301 participants entered the website. Of those, 13 were dis-
qualified because they did not agree to participate (n = 2), did not have a
3- to 5-year-old child (n = 10), or did not live with the child during most
of the week (n = 1). Another 51 entries were deleted from the database: 11
were duplicates (assessed by checking email and IP addresses) and 40 par-
ticipants did not complete the 36-item Children’s Behavior Questionnaire
scale. The final data set contained 237 participants, mostly female caregiv-
ers (93.2%). Children’s mean age was 3.9 yr. (SD = 0.8), and sex of the child
was almost equally divided (boys: 53.6%, girls: 46.4%). The study was re-
viewed and approved by the Institutional Review Board of the Baylor Col-
lege of Medicine; all caregivers completed online informed consent prior
to data collection.
Measures
Background characteristics.—Parents or guardians were asked to report
some family demographics in the beginning of the online survey includ-
ing child’s age, sex, and their relationship to the child. The participant’s
sex, race, living situation, highest educational level attained, employment,
and annual household income were assessed at the end of the survey.
Temperament scales.—The 36-item scale of the Children’s Behavior
Questionnaire (Putnam & Rothbart, 2006) was used to assess the three
broad factors of temperament: Surgency, Negative Affectivity, and Effort-
ful Control, consisting of 12 statements for each trait. Parents were asked
to report using a 7-point scale, with anchors 1: Extremely untrue of your
child and 7: Extremely true of your child, on how well each statement de-
scribed their child’s reaction to a given situation within the past 6 months;
 Validation of CBQ and One-Item Temperament Scale 201

a “not applicable” option was provided if parents could not answer be-
cause they had not seen their child in the situation described. The Chil-
dren’s Behavior Questionnaire higher order factors have been replicated
across multiple age samples among different countries and showed ad-
equate internal consistency reliability ranging from .61 to .94 (e.g., Aha-
di, et al., 1993; Kochanska, De Vet, Goldman, Murray, & Putnam, 1994;
Rothbart, et al., 2001; Majdandžić & Van den Boom, 2007; Sleddens, et al.,
2011). The very short scale (Putnam & Rothbart, 2006; Hughes, et al., 2008;
Ward, et al., 2008; O’Connor, et al., 2010; Sleddens, et al., 2011) also showed
adequate internal consistency (Cronbach’s alphas ranging from .63 to .76
for Surgency, from .65 to .79 for Negative Affectivity, and from .67 to .74
for Effortful Control). Convergent validity included parental agreement
(Rothbart, et al., 2001; Majdandžić & Van den Boom, 2007; Majdandžić, Van
den Boom, & Heesbeen, 2008) and prediction of social (Rothbart, Ahadi,
& Hershey, 1994; Rotbhart, et al., 2001) and laboratory behavior patterns
(Kochanska, Murray, Jacques, Koenig, & Vandegeest, 1996; Majdandžić &
Van den Boom, 2007; Majdandžić, et al., 2008).
One-item scale.—The one-item temperament scale included three re-
sponse options reflecting the three higher order temperament factors (Sur-
gency, Negative Affectivity, and Effortful Control) from the Children’s
Behavior Questionnaire. For a description of the one-item scale and op-
erational definitions, see Table 1. This one-item scale was developed by
two of the authors (SH and TO) and agreed upon by the whole group. The
two co-authors (SH and TO), one an expert in child development (SH) and
the other a pediatrician (TO), wrote single item statements that reflected

Table 1
Single-item Temperament Scale
Please read the statements below and select which one best
describes your 3- to 5-year-old child (select just the best one)
The “Surgency/Extraversion” child
This child has lots of energy, is easily excited, and often goes fast on the playground. This
child enjoys meeting new people and going to new places.
The “Negative Affectivity” child
This child often shows his frustration or discomfort, and easily becomes sad when not
able to finish a project. This child is often afraid of the dark, and when upset may be dif-
ficult to calm down.
The “Effortful Control” child
This child likes to listen to rhymes and songs. When working on a project, this child can
concentrate deeply, and carefully follows rules and instructions. When something chang-
es, this child quickly notices.
Note.—Operational definition of the three higher order temperament factors: Surgency (ten-
dency to perform impulsive and active behavior), Negative Affectivity (predisposition to ex-
perience negative affective states), and Effortful Control (the ability of a child to control at-
tentional processes and behavior).
202 E. F. C. SLEdDens, et al.

the general theory and the items (Rothbart, et al., 2001; Putnam & Roth-
bart, 2006) for each of the three major child temperament factors. Each vi-
gnette attempted to capture the original subscales defined by the higher
order factor (Rothbart, et al., 2001): Surgency by impulsivity, lack of shy-
ness, activity level, and high intensity pleasure; Negative Affectivity by
anger, discomfort, sadness, difficulty to sooth, and fear; and Effortful Con-
trol by inhibitory control, attentional focusing, low intensity pleasure, and
perceptual sensitivity. Each statement was reviewed and revised by the
co-authors until consensus was reached about optimal wording. A single
item was selected to identify the dominant temperament characteristic of
the child in order to simplify the task of tailoring an intervention to the
child’s temperament.
Statistical Analyses
Classical test theory item analysis, 36-item temperament scale.—Rele-
vant scale items of the 36-item scale were combined to create mean fac-
tor scores for Surgency, Negative Affectivity, and Effortful Control. Items
marked “does not apply” were not included in the scores. Missing data
were not present because the participants were forced to answer all of the
questions or the computer would not progress. Internal consistency reli-
ability was evaluated by calculating both Cronbach’s alpha and average
corrected item-total correlations. The minimum acceptable cutoff point for
Cronbach’s alphas was .70, and for item-scale correlations cutoff points
were between .15 and .30 (Nunnally & Bernstein, 1994). Correlations of .30
or above were considered good, and below .15 were considered unreliable
since they would indicate lack of homogeneity of the items within a scale.
Principal factor analysis, 36-item temperament scale.—Principal factor
analysis (exploratory in nature) with oblique rotation yielded factor load-
ings for each item as well as the percent variance explained by each factor.
In accordance with the original validation studies (Rothbart, et al., 2001;
Putnam & Rothbart, 2006), a factor loading of .25 or higher was an indi-
cation of a reasonably high factor loading. Items with no numerical re-
sponse were imputed with the mean score of the factor to which the item
belonged. On average, respondents chose the “not applicable” option less
than 1% of the time. The frequency of “not applicable” responses was
identified for each item. Item 6 (“Prepares for trips and outings by plan-
ning things he/she will need”) had the largest number of “not applicable”
answers (n = 12, 5.1%). To verify the robustness of the results, a principal
factor analysis was performed on only complete cases (n = 183), excluding
54 respondents with missing values. Parallel analysis was performed us-
ing the SPSS syntax (O’Connor, 2000) to determine accurately the number
of factors to retain in the factor analysis. The program was forced to gen-
erate 1,000 samples of random number matrices and corresponding ei-
 Validation of CBQ and One-Item Temperament Scale 203

genvalues, each of which was based on 183 cases and 36 variables. The
eigenvalues resulting from this procedure were compared with the eigen-
values found following principal factor analysis on the data. Factors were
retained when eigenvalues from the actual data (following principal fac-
tor analysis) were greater than the eigenvalues from the random data.
Item response modeling analyses, 36-item temperament scale.—Rasch mod-
eling (partial credit model) was performed on all cases using the “Con-
Quest” software program (Wu, Adams, & Haldane, 2007) which provided
item parameter difficulty estimates, item fit statistics, Wright maps, and
reliability indices. Item difficulty refers to the item’s location on the un-
derlying temperament trait, and reflects how difficult it was for a per-
son to respond affirmatively to an item (higher values = more difficult).
Item fit was determined by computing weighted mean square fit statis-
tics for each item, which indicated whether residuals varied as much as
expected given the observed distribution. Items with a weighted fit sta-
tistic < .75 or > 1.33, and for which the weighted t statistic was < −2.00 or
> 2.00, were considered to fit poorly (Adams & Khoo, 1996). The Wright
map visually linked the distribution of respondents (indicated by Xs on
the left side of the Wright map) on the latent temperament factors to the
distribution of individual item difficulties (represented on the right side
by item number), using the same metric, i.e., a logit scale centered at a
mean of zero. The items and respondents should be located between −3
and +3 logits. Also included in the Wright map is the location of the items
by threshold (the 7-point Likert-type scale is separated by six threshold
points, where Threshold 1 refers to the threshold between response op-
tion 1: Extremely untrue of your child and response option 2: Quite un-
true of your child. Item separation reliability, which indicated “how well
the sample of subjects had spread the items along the measure of the test”
(Fisher, 1992, p. 238), were calculated for the three factors. Finally, visual
examination of the item response functions evaluated the functioning of
the 7-point response format.
Analysis of variance and Pearson chi-square analyses of the one-item tem-
perament scale.—The average scores of the three temperament factors (Sur-
gency, Negative Affectivity, Effortful Control) were compared for each of
the three temperament groups as defined by the one-item temperament
scale. This was conducted by means of a three-level group ANOVA (Sur-
gency, Negative Affectivity, Effortful Control), followed by Bonferroni and
Tukey’s HSD post hoc analyses, adjusting p values for multiple testing to
detect differences in mean scores on each of the three temperament fac-
tors between the three groups. Pearson chi-square analyses with categori-
cal variables were used with a 3 × 3 contingency table comparing tempera-
ment characteristics of the children, defined by parents using the one-item
204 E. F. C. SLEdDens, et al.

scale, and the 36-item scale. Mean scores were calculated for the three tem-
perament factors of Surgency, Negative Affectivity, and Effortful Control,
and a child was characterized by a temperament characteristic based on
the highest mean score across the three factors.
Results
The study sample was ethnically diverse (Table 2). The majority were
White (39.2%), but Hispanics (25.3%) and African-Americans (23.6%) were
also represented. A minority of the participants were combined into “oth-
er,” consisting of American Indians, Native Hawaiians, Pacific Islanders,
and Asians (10.5%). In total, 209 participants (88.2%) were from Texas
(United States). The other 28 participants reported living in other states
within the United States (n = 24), Canada (n = 1), or these data were missing
(n = 3). Most participants indicated they were married or lived with a signif-
icant other (75.5%). A majority reported higher levels of education (59.1%
indicated having a college degree or higher) and high annual household
income (46.0% indicated a household income above $60,000 a year).
Psychometric Properties of the 36-item Temperament Scale
Classical test theory item analysis.—Internal consistency (Cronbach’s al-
pha) ranged from .69 to .78 (Table 3). The average item-total correlations
among the three factors suggested good consistency of the items (Surgen-
cy = .41, Negative Affectivity = .37, Effortful Control = .33). For the Surgen-
cy factor, the corrected item-total correlations of all items were above the
value of .30. For the two other higher order factors, some items fell below
the value of .30, but still above the critical cut-off point of .15 (Negative Af-
fectivity = two items; Effortful Control = five items). One Effortful Control
item had an item-total correlation (after correlation for overlap) of .05, far
below the critical value of .15. Cronbach’s alpha increased from .69 to .71
when this item was removed.
Principal factor analysis.—Principal factor analysis revealed 12 factors
having eigenvalues greater than 1.00. Forcing a three-factor structure, as
proposed by Rothbart, et al. (2001) and Putnam and Rothbart (2006), re-
vealed the three higher order factors Surgency, Negative Affectivity, and
Effortful Control (Table 3). Results of the parallel procedure (eigenvalues
from random data were lower than the eigenvalues of the three factors
found following PFA) supported the retention of the three-factor struc-
ture. The three-factor solution accounted for 25.1% of the total variance.
Most of the variance was explained by the Surgency factor (9.4%) with Ef-
fortful Control explaining 8.4% and Negative Affectivity explaining 7.3%,
respectively. Item 30 on Effortful Control (“My child approaches places
he/she has been told are dangerous slowly and cautiously”) had a factor
loading below .25 (factor loading of .09). Factor analyses on the complete
 Validation of CBQ and One-Item Temperament Scale 205

Table 2
Participant Characteristics (N = 237)
Description n %

Sex
Male 16 6.8
Female 221 93.2
Relationship to child
Parent 227 95.8
Grandmother 5 2.1
Aunt 4 1.7
Female guardian 1 0.4
Race
Black or African-American 56 23.6
White or Euro-American 93 39.2
Hispanic 60 25.3
Other 25 10.5
No response 3 1.3
Living situation
Married or living with a significant other 179 75.5
Single, never married 21 8.9
Divorced, separated or widowed 35 14.8
No response 2 0.8
Education
Attended some high school 12 0.8
High school graduate or GED 17 7.2
Technical school 12 5.1
Some college 64 27.0
College graduate 72 30.4
Post-graduate study 68 28.7
No response 2 0.8
Employed
Yes 160 67.5
No 75 31.6
No response 2 0.8
Annual household income
Less than $10,000 10 4.2
$10,000–$19,999 15 6.3
$20,000–$39,999 45 19.0
$40,000–$59,999 56 23.6
$60,000 or more 109 46.0
No response 2 0.8

cases (n = 183) obtained similar results (data not reported in table), and the
factor loading of Item 30 only slightly improved from .09 to .11.
Item response modeling analyses.—All items of the three higher order
factors (Surgency, Negative Affectivity, and Effortful Control) had both
weighted mean square statistic and t statistic within the acceptable rang-
es (Table 3). Fig. 1 presents the item-respondent Wright map for the three
 Table 3

206
Psychometric Properties for the Items of the 36-item Scale of the Children’s Behavior
Questionnaire Clustered by the Three Higher Order Temperament Factors
No. 36-item Scale Children’s Behavior Questionnaire M SD CITC Factor Est. SE MNSQ t
Surgency/Extraversion, Factor 1: Cronbach’s alpha = .78 4.59 0.86
25 Is full of energy, even in the evening 5.80 1.23 0.41 0.47 −0.69 0.04 0.97 −0.2
4 Likes going down high slides or other adventurous activities 5.56 1.56 0.34 0.473 −0.37 0.03 1.08 0.8
16 Likes to go high and fast when pushed on a swing 5.27 1.60 0.38 0.46 −0.23 0.03 1.05 0.6
28 Likes rough and rowdy games 5.03 1.68 0.50 0.58 −0.13 0.03 0.94 −0.6
13 Prefers quiet activities to active games (R) 4.99 1.29 0.38 0.46 −0.11 0.04 0.99 −0.1
31 Is slow and unhurried in deciding what to do next (R) 4.38 1.51 0.37 0.45 0.02 0.03 1.03 0.4

E. F. C. SLEdDens, et al.
10 Seems to be at ease with almost any person 4.42 1.84 0.44 0.46 0.09 0.03 1.05 0.7
19 Takes a long time in approaching new situations (R) 4.33 1.49 0.49 0.542 0.10 0.03 0.94 −0.7
22 Is sometimes shy even around people (s)he has known a long
time (R) 4.14 1.94 0.41 0.42 0.16 0.03 1.09 1.1
1 Seems always in a big hurry to get from one place to another 3.94 1.68 0.41 0.49 0.28 0.03 1.03 0.4
7 Often rushes into new situations 3.86 1.54 0.50 0.603 0.39 0.03 0.93 −0.9
34 Sometimes turns away shyly from new acquaintances (R) 3.37 1.67 0.36 0.363 0.49 0.11 1.08 0.9
Negative Affectivity, Factor 2: Cronbach’s alpha = .74 4.13 0.85
2 Gets quite frustrated when prevented from doing something
(s)he wants to do 5.41 1.38 0.31 0.431 −0.51 0.03 1.04 0.4
32 Gets angry when (s)he can’t find something (s)he wants to play
with 4.41 1.56 0.50 0.59 −0.13 0.03 0.89 − 1.4
5 Is quite upset by a little cut or bruise 4.14 1.73 0.47 0.52 −0.06 0.03 0.92 −1.1
35 Becomes upset when loved relatives or friends are getting ready
to leave following a visit 4.48 1.81 0.28 0.28 −0.08 0.10 1.15 1.8
8 Tends to become sad if the family’s plans don’t work out 4.40 1.48 0.45 0.53 −0.06 0.03 0.91 −1.0
11 Is afraid of burglars or the ‘boogie man’ 4.20 1.84 0.30 0.34 −0.03 0.03 1.08 1.0
26 Is not afraid of the dark (R) 3.69 1.93 0.26 0.35 −0.01 0.03 1.11 1.4
20 Hardly ever complains when ill with a cold (R) 3.70 1.64 0.35 0.42 0.09 0.03 0.99 −0.1
29 Is not very upset at minor cuts or bruises (R) 3.89 1.63 0.37 0.40 0.12 0.03 1.02 0.3
23 Is very difficult to soothe when (s)he has become upset 3.58 1.77 0.34 0.45 0.16 0.03 1.03 0.4
(continued on next page)
 Table 3 (cont’d)
Psychometric Properties for the Items of the 36-item Scale of the Children’s Behavior
Questionnaire Clustered by the Three Higher Order Temperament Factors

Validation of CBQ and One-Item Temperament Scale

No. 36-item Scale Children’s Behavior Questionnaire M SD CITC Factor Est. SE MNSQ t
Negative Affectivity, Factor 2: Cronbach’s alpha = .74 4.13 0.85
14 When angry about something, (s)he tends to stay upset for ten
minutes or longer 3.59 1.73 0.46 0.54 0.25 0.03 0.95 −0.7
17 Seems to feel depressed when unable to accomplish some task 3.70 1.60 0.41 0.49 0.27 0.03 1.00 0.1
Effortful Control, Factor 3: Cronbach’s alpha = .69 5.26 0.71
21 Likes the sound of words, such as nursery rhymes 5.96 1.11 0.29 0.31 −0.26 0.04 1.00 0.1
24 Is quickly aware of some new item in the living room 5.74 1.37 0.44 0.55 −0.24 0.04 0.92 −0.7
27 Sometimes becomes absorbed in a picture book and looks at it
for a long time 4.92 1.40 0.26 0.27 −0.17 0.04 1.03 0.4
9 Likes being sung to 5.60 1.41 0.23 0.29 −0.10 0.04 1.07 0.6
3 When drawing or coloring in a book, shows strong concentration 5.43 1.52 0.43 0.52 −0.08 0.03 0.93 −0.8
36 Comments when a parent has changed his/her appearance 5.54 1.44 0.42 0.57 −0.06 0.12 0.95 −0.4
18 Is good at following directions 5.31 1.29 0.40 0.532 −0.01 0.04 0.94 −0.5
15 When building or putting something together, becomes very in-
volved in what (s)he is doing, and works for long periods 5.13 1.50 0.41 0.49 0.02 0.03 0.94 −0.6
12 Notices it when parents are wearing new clothing 5.15 1.81 0.42 0.59 0.15 0.03 1.00 0
33 Enjoys gentle rhythmic activities such as rocking or swaying 5.01 1.60 0.28 0.28 0.16 0.03 1.07 0.8
30 Approaches places (s)he has been told are dangerous slowly and
cautiously 4.79 1.60 0.05 0.09 0.28 0.03 1.22 2.3
6 Prepares for trips and outings by planning things (s)he will need 4.45 1.63 0.28 0.38 0.31 0.03 1.06 0.7
Note.—N = 237; CITC: corrected item-total correlation; R: reversed coding; mean score ranges for Surgency 2.33–6.92, Negative Affectivity 1.25–
6.17, and Effortful Control 2.08–6.92; Cronbach’s alpha Surgency .78 (n = 221), Negative Affectivity .74 (n = 213), Effortful Control .69 (n = 209);
EAP/PV reliability .77, .75, .69 for Surgency, Negative Affectivity and Effortful Control respectively; values of items indicated in bold are critical;
1, 2, 3
Principal factor analysis loadings greater than or equal to .25 (for items loading on another factor than intended), Factor 1 (Surgency), Factor
2 (Negative Affectivity), Factor 3 (Effortful Control); Item response modeling item difficulty estimate: Est. (SE), presented in ascending order of
difficulty; Item response modeling MNSQ (t) corresponds to weighted mean square statistics and weighted t statistics: weighted mean square
statistics < .75 and > 1.33 and t statistics < −2.0 and > 2.0 are both outside acceptable range.

207
208 E. F. C. SLEdDens, et al.

higher order temperament factors (Surgency: Fig. 1a, Negative Affectiv-

ity: Fig. 1b, Effortful Control: Fig. 1c). The distribution of items indicated
no item difficulty estimates covered the scores that were extremely easy or
extremely difficult (the logits ranged from −0.69 to 0.49). The best distribu-
tional results were for Surgency (Fig. 1a). However, the upper end of the
continuum remained uncovered by items with higher difficulty. The dis-
tribution of Negative Affectivity items (Fig. 1b) was clustered at the mid-
dle: the 12 items did not adequately distinguish among parents with chil-
dren rated high and low on Negative Affectivity, and several items had
overlapping difficulty estimates. Effortful Control had an especially poor
distribution of items (Fig. 1c). The items covered a restricted portion of the
range of respondents’ answers, suggesting inadequate content validity for
children at the more-difficult-to-agree end of the scale. Several items occu-
pied the same location on the Effortful Control latent variable, indicating
they had the same difficulty and could be removed from the scale without
affecting content validity.
For all three latent temperament variables, every item had the first
threshold targeting none of the respondents, suggesting the “extremely
untrue of your child” response option was not chosen by the respond-
ents. This indicated that the scale had skewed content representation of
the construct and Threshold 1 could be removed. Moreover, the second
threshold targeted an Effortful Control of −0.2 and less, but only approxi-
mately four respondents had this score, suggesting also that the second
threshold could be eliminated. For all 12 items of Negative Affectivity,
the sixth threshold targeted none of the respondents, indicating that the
response option “extremely true of your child” was not chosen by the re-
spondents. The item separation reliability estimates were .77, .75, and .69
for Surgency, Negative Affectivity, and Effortful Control, respectively. Fi-
nally, the functioning of the 7-point response scale was assessed by visual
examinations of the item response functions. The pattern of response op-
tions suggested that several items functioned more like a 5-point rather
than a 7-point response format.
Validation of a One-item Temperament Scale using the 36-item Tempera-
ment Scale.—ANOVAs across the three temperament groups defined by
the one-item temperament scale (Surgency, n = 108; Negative Affectivity,
n = 30; Effortful Control, n = 99) indicated significant differences among the
means on the three temperament factors (Surgency: F2,236 = 31.35, p < .001,
η2 = 0.46; Negative Affectivity: F2,236 = 8.01, p < .001, η2 = 0.25; Effortful Con-
trol: F2,236 = 8.26, p < .001, η2 = 0.26).
Bonferroni post hoc tests revealed that parents who selected the “Sur-
gency” response options had higher scores on the Surgency factor (using
the 12-item scale; M = 5.02, SD = 0.72) compared to the scores on this fac-
tor for parents selecting the other options (Negative Affectivity: M = 4.30,
 Validation of CBQ and One-Item Temperament Scale 209
Logits Resp Items *Items by Threshold Logits Resp Items *Items by Threshold
--------------------------------------------------- ----------------------------------------------------
| |7.6 | |8.6 14.6 17.6 20.6 29.6
| |19.6 | |32.6
| |31.6 34.6 | |5.6 23.6 35.6
| | | |
| | | |
| |1.6 | |
| | | |
| |13.6 1 | |
| | | |11.6
X| |10.6 | |2.6 17.5 26.6
X| | | |
| | | |14.5
X| |22.6 | |29.5
X| | | |20.5
X| |7.5 28.6 X| |8.5 23.5
1 X| |34.5 X| |
XX| |1.5 XX| |5.5 32.5
XX| |16.6 XX| |
XX| |19.5 XX| |
XXX| | XXX| |26.5
XXX| |25.6 31.5 XXXX| |11.5
XXXX| | XXXX| |
XXXXX| |34.4 XXXXXX| |
XXXX| |4.6 XXXXX| |17.4
XXXXX| |7.4 13.5 22.5 XXXXX|17 |20.4 23.4 35.5
XXXXXXX|34 |1.4 10.5 34.3 XXXX|14 |14.4
XXXXXXXXX| | XXXXXX| |
XXXXXXXXXX|7 | XXXXXXXX|23 |14.3 23.3 29.4
XXXXXXXXX| |28.5 XXXXXXXXXX|20 29 |20.3 26.4
XXXXXXXXX|1 |19.4 22.4 31.4 XXXXXXXX| |5.4 11.4
XXXXXXXXX| |22.3 0 XXXXXXXX|26 |
XXXXXXXXX|22 |10.4 16.5 XXXXXXX|11 |26.3 29.3
XXXXXXXXXX|10 19 | XXXXXXX|5 8 35 |2.5 8.4 17.3 32.4 35.4
XXXXXXXXXX| |7.3 XXXXXXXXX|32 |
0 XXXXXXXXXX|31 |1.3 4.5 10.3 XXXXXXXXXX| |
XXXXXXX| |13.4 XXXXXXX| |5.3 11.3
XXXXXXX|13 | XXXXX| |23.2 35.3
XXX|28 |19.3 28.4 31.3 34.2 XXX| |14.2 26.2
XXXX|16 |25.5 XXX| |32.3
XXXXX| |16.4 XXX| |
XXX| |1.2 22.2 28.3 XXX| |5.2 8.3 11.2
XX|4 |4.4 10.2 | |17.2 20.2 29.2
XX| |16.3 X|2 |
X| |4.3 7.2 XX| |35.2
X| | X| |
X| |25.4 28.2 X| |2.4
X| | | |
X|25 | X| |2.3
| |4.2 16.2 | |
X| |13.3 25.3 | |32.2
| |19.2 31.2 34.1 | |8.2
| | | |2.2
| |22.1 | |
-1 | |1.1 -1 | |
| | | |14.1
| |7.1 10.1 16.1 28.1 | |11.1 17.1 35.1
| |25.2 | |
| |4.1 | |23.1 26.1
| | | |
| |13.1 13.2 19.1 25.1 31.1 | |2.1 5.1 8.1 20.1 29.1 32.1
==================================================== ======================================================

Fig. 1a. Item-respondent map for Surgen- Fig. 1b. Item-respondent map for Negative
cy. N = 237; each “X” represents 1.4 cases; Affectivity. N = 237; each “X” represents 1.6
*items by thresholds show the item first cases; *items by thresholds show the item
followed by the threshold (e.g., 7.4 refers to first followed by the threshold (e.g., 8.4 re-
Item 7 and Threshold 4). fers to Item 8 and Threshold 4).
210 E. F. C. SLEdDens, et al.

Logits Resp Items *Items by Threshold

----------------------------------------------------
| |
| |6.6
| |18.6 30.6
| |27.6
| |
| |
| |
| |33.6
| |
X| |15.6
| |
| |
X| |
| |
XX| |
XX| |
X| |3.6 12.6
1 XX| |9.6 36.6
XX| |
X| |
XX| |6.5
XXX| |21.6
XXXXXXX| |30.5
XXX| |24.6
XXX| |27.5
XXX| |
XXXXX| |
XXXXXX| |
XXXXXXXXXX| |15.5 33.5
XXXXXXXX| |
XXXXXXXXX| |18.5
XXXXXX| |6.4 12.5
XXXXXXX| |
XXXXXXXX|6 |
XXXXXXXX|30 |3.5
XXXXXXX| |36.5
XXXXXX| |9.5 24.5
XXX|12 33 |27.4
XXX| |12.4 30.4 33.4
XXX| |6.3
0 XXX|15 18 |12.3 15.4
XX| |3.4 30.3
X|3 9 36 |21.5
X| |15.3 27.3 33.3 36.4
X|27 |3.3 9.4 12.2 18.4
X| |
|21 24 |6.2 24.4
| |30.2
| |18.3 36.3
| |9.3 24.3 33.2
| |9.2
| |21.4
| |12.1
| |36.2
| |3.2 15.2 24.2
| |21.3 30.1
| |21.2
| |18.2 33.1 36.1
| |21.1
| |3.1 6.1 9.1 15.1 18.1 24.1 27.1 27.2
===============================================================

Fig. 1c. Item-respondent map for Effortful Control. N = 237; each “X” represents 1.8 cases;
*items by thresholds show the item first followed by the threshold (e.g., 9.4 refers to Item 9
and Threshold 4).
 Validation of CBQ and One-Item Temperament Scale 211

Table 4
Average Scores on the Three Factors of the 36-item Temperament
Scale for Parents Classifying their Children on the One-item
Scale and p Values For Pairwise Between-Groups Comparisons
36-item Scale One-item Temperament Scale (1–2) (1–3) (2–3)
(3 Factors)
Surgency Negativ Effortful
(n = 108) (1) Affectivity Control (n = 99)
(n = 30) (2) (3)
M SD M SD M SD
Surgency 5.02 0.72 4.30 0.96 4.22 0.75  < .001*  < .001* 1.0
Negative Affectivity 4.08 0.83 4.68 0.78 4.00 0.84 .002* 1.0  < .001*
Effortful Control 5.18 0.68 4.92 0.72 5.45 0.70 .209 .014* .001*
Note.—N = 237; using Bonferroni correction for adjusting post hoc multiple comparisons;
statistical significance was assigned at the p < .05 level; *p value remained significant using
Tukey α, 36-item Temperament Scale with answer options ranging from 1 to 7.

SD = 0.96; Effortful Control: M = 4.22, SD = 0.75; ps < .001). No significant

differences in mean scores on Surgency were present between parents
classifying their child as having either a “Negative Affectivity” or an “Ef-
fortful Control” temperament. Similar findings were obtained for the oth-
er two temperament factors (Negative Affectivity and Effortful Control;
Table 4). When using Tukey α for adjusting post hoc multiple comparisons
to identify significant mean differences, similar results were found (data
not reported).
Three participants had similar high mean scores on two of the three
factors of the 36-item temperament scale and these were deleted because
they could not be properly categorized into a single temperament trait
(Table 5). The Pearson chi-squared statistic was 70.79 (p value of < .001),
suggesting an association between the temperament constructs as meas-
Table 5
Parental Characterization of Child Temperament Based
on the One-item and 36-item Temperament Scale
36-item Scale One-item Temperament Scale Total Sensitivity
(%)
Surgency Negative Effortful
Affectivity Control
Surgency 43 4 9 56 76.79
Negative Affectivity 6 13 5 24 54.17
Effortful Control 56 13 85 154 55.19
Total 105 30 99 234
Note.—n = 234; a respondents with similar mean scores (and the highest mean scores) on two
of the three Children’s Behavior Questionnaire higher order factors was excluded from the
analyses (n = 3); χ2 = 70.79, df = 4, p < .001; statistical significance were assigned at the p < .05 lev-
el; sensitivity was calculated to identify the percentage of caregivers identifying their child
on a particular temperament trait with both temperament scales (36-item scale is the “gold-
en standard”).
212 E. F. C. SLEdDens, et al.

ured by the two methods. A sensitivity score was calculated to identify the
percentage of caregivers identifying their child on a particular tempera-
ment trait with both temperament scales (36-item scale is the “gold stand-
ard”). Thus, it is the probability a parent correctly identifies their child on
a particular temperament trait. This score ranged from 54.17% for Nega-
tive Affectivity to 76.79% for Surgency. However, a substantial number of
participants (n = 56) characterized their child as “Surgency” on the one-
item temperament scale, but the highest mean score was for the Effortful
Control factor (Table 5).
Secondary analyses showed that for the group of children whose
parents chose “Surgency” on the one-item scale, the means of the three
temperament factors were: Surgency 4.71 (SD = 0.56), Negative Affectiv-
ity 4.04 (SD = 0.77), and Effortful Control 5.61 (SD = 0.44). Pearson’s corre-
lation analysis showed that Surgency and Effortful Control were related
(r = .35, p < .01), whereas the correlations between Effortful Control-Nega-
tive Affectivity and Surgency–Negative Affectivity were not statistically
significant. Thirteen participants identified their child’s temperament as
“Negative Affectivity” on both scales. Thirteen other participants also de-
scribed their child’s temperament as being “Negative Affectivity” using
the one-item scale, but had a higher mean score on Effortful Control using
the 36-item Children’s Behavior Questionnaire scale (Table 5). Secondary
analyses showed that the means for this group of children for the three
temperament factors were as follows: Surgency 3.76 (SD = 0.68), Negative
Affectivity 4.29 (SD = 0.81), and Effortful Control 5.45 (SD = 0.62) on the
one-item scale. Findings of the correlation analysis showed that none of
the temperament factors were significantly correlated.
Discussion
The psychometric properties of the very short scale of the Children’s
Behavior Questionnaire, using classical test theory, principal factor analy-
sis, and item response modeling, were found to be acceptable to outstand-
ing. These methods complement each other and are needed for a thorough
psychometric evaluation of the temperament scale. The factor analytic
results closely resembled those of the original United States (Putnam &
Rothbart, 2006) and Dutch studies (Sleddens, et al., 2011), after forcing the
principal factor analysis to retrieve three factors. Most of the variance in
the present study was explained by Surgency, in contrast to the study of
Putnam and Rothbart (2006) and the Dutch study of Sleddens, et al. (2011),
where Negative Affectivity was the factor accounting for the highest vari-
ance. Factor loadings in the present study were slightly higher than in the
Dutch study, which may have been due to the Children’s Behavior Ques-
tionnaire in this study being administered in isolation from the longer
scales of the Children’s Behavior Questionnaire, or to differences in cul-
 Validation of CBQ and One-Item Temperament Scale 213

ture. Better psychometric properties might be expected when solely ad-

ministering the 36 items, because respondents may be more inclined to be
conscientious when completing a shorter scale (Putnam & Rothbart, 2006).
The classical test theory results indicated that internal reliability es-
timates were acceptable and most items discriminated well. However,
Item 30 (“My child approaches places he/she has been told are dangerous
slowly and cautiously”) had a corrected item-total correlation of .05 which
is far below the critical cutoff point of .15 (Nunnally & Bernstein, 1994).
In addition, this item had a low factor loading following principal factor
analysis (.09), but the item did not have a significant loading on one of the
two other higher order factors. These results are not consistent with other
studies (Putnam & Rothbart, 2006; Sleddens, et al., 2011), and must be rep-
licated before definitive conclusions can be drawn. When comparing the
internal reliability estimates of the current study to other studies in which
the 36-item scale was administered in isolation from the longer versions
(Hughes, et al., 2008; O’Connor, et al., 2010; Ward, et al., 2008), slight devia-
tions were found. In this study, Cronbach’s alpha for Surgency was higher
compared to the others. Cronbach’s alpha for Effortful Control was high-
er in the studies of Hughes, et al. (2008) and O’Connor, et al. (2010); both
studies reported samples of predominantly African-American and His-
panic Head Start preschoolers. Ward, et al. (2008) found a slightly higher
internal reliability coefficient for Negative Affectivity in a small group of
preschoolers.
The item response modeling analyses provided a complementary in-
depth assessment of the psychometric properties of the very short scale of
the Children’s Behavior Questionnaire. Although fit indices were within
acceptable ranges and reliability was adequate, the findings of the Wright
map were less optimistic. The distribution of participants along the con-
tinuum appeared to be normal; however, the items were not well distrib-
uted along the latent variables, and no item difficulty estimates covered
the scores that were extremely easy or difficult. Especially for the Effort-
ful Control latent variable, only a restricted portion of the distribution was
covered by items with lower difficulty, suggesting inadequate content va-
lidity. Moreover, the 7-point response format did not work well for several
items. Future research needs to use the item response modeling approach
on the Children’s Behavior Questionnaire using different samples includ-
ing countries outside the United States, and the longer scales of the Chil-
dren’s Behavior Questionnaire should be evaluated to identify items that
enhance the distributional properties of the scale, and remove redundant
items. This is the first study reporting the factorial validity of the 36-item
Children’s Behavior Questionnaire scale when administered in isolation
from the longer versions. Moreover, this study is the first using item re-
sponse modeling approaches for analyzing the Children’s Behavior Ques-
214 E. F. C. SLEdDens, et al.

tionnaire, beyond more classical test theory approaches for psychometric

evaluation.
The newly developed one-item temperament scale is a reasonably
valid alternative measure to use in studies where brief and quick assess-
ment of child temperament is desired, such as tailoring interventions. The
vignettes contained the information to correctly identify child tempera-
ment (sensitivity percentages ranging from 54 to 77%), despite some vio-
lations of the hypothesis, i.e., a large group of parents characterizing their
child’s temperament as Surgency or Negative Affectivity on the one-item
scale, having higher mean scores on one of the other temperament factors
using the 36-item scale of the Children’s Behavior Questionnaire. Such a
brief, single-item scale is not intended to replace the multi-dimensional
scales for assessing child temperament for observational studies, but rath-
er will allow scientists to translate the findings from decades of research
on child temperament to develop novel interventions tailored according
to the child’s behavioral characteristics that parents perceive to most ex-
emplify their child’s temperament. Brief vignettes, as used in this study,
were more effective when including fewer details or texts for participants
to attend to (Sha & Pan, 2009). Therefore, when there are time constraints
and when a large battery of questionnaires needs to be administered, the
one-item scale would be an appropriate alternative to assist in tailoring
interventions.
The strengths of this study include the in-depth assessment of psy-
chometric properties of the Children’s Behavior Questionnaire very short
scale administered in isolation from the longer scales, and using item re-
sponse modeling approaches in addition to more traditional approaches.
Moreover, the sample was ethnically diverse. The weaknesses include a
potential sample bias since a large group of participants reported a house-
hold income of $60,000 or more a year (46%). Putnam and Rothbart (2006),
however, reported that internal consistency for low-income participants
was only lower for the 94-item short scale, not the very short scale. The
majority of participants lived in Texas, which lowered generalizability to
other populations. Another limitation is the loss of some important infor-
mation when the very short scale, and more specifically the one-item tem-
perament scale, was administered instead of one of the more precise lon-
ger scales of the Children’s Behavior Questionnaire. However, given the
findings of this study, these short scales are appropriate to use in case of
time constraints and to reduce participant burden.
Implications for Research and Practice
The current study showed adequate psychometric properties of the
very short scale Children’s Behavior Questionnaire among a sample of
United States parents of 3- to 5-year-old children. Item response model-
 Validation of CBQ and One-Item Temperament Scale 215

ing indicated lack of item coverage among respondents on the extremes

of the latent temperament variables. Future cross-cultural research is nec-
essary, applying item response modeling to other samples that complet-
ed the 36-item scale of the Children’s Behavior Questionnaire. Item re-
sponse modeling approaches should also be applied to longer scales of
the Children’s Behavior Questionnaire to generate additional items at dif-
ferent levels of difficulty of the temperament variables to improve assess-
ment of the full range of these constructs with non-redundant items. Ef-
forts to use these more advanced statistical techniques are upcoming in
health behavior research, and needed for proper development and valida-
tion of questionnaires. The newly developed one-item temperament scale
has promising results (i.e., moderate predictive value) and is potentially
useful in intervention studies tailoring to child temperament to decrease
participant burden.
References
Adams, R. J., & Khoo, S. T. (1996) Quest-Interactive test analysis system. Victoria, Aus-
tralia: The Australian Council for Education Research.
Ahadi, S. A., Rothbart, M. K., & Ye, R. (1993) Children’s temperament in the U.S.
and China: similarities and differences. European Journal of Personality, 7, 359-377.
Baranowski, T., O’Connor, T., Hughes, S., Beltran, A., Baranowski, J., Nicklas, T.,
Sleddens, E., Thompson, D., Lu, A., & Buday, R. (in press) Smart phone video
game simulation of parent-child interaction: learning skills for effective vegetable
parenting. In S. Arnab, I. Dunwell, & K. Debattista (Eds.), Serious games for health-
care: applications and implications. Hershey, PA: IGI Global.
Beltran, A., O’Connor, T. M., Hughes, S. O., Baranowski, J., Nicklas, T. A., Thompson,
D., & Baranowski, T. (in press) Alpha test of a video game to train parents in
effective vegetable parenting. The Games for Health Journal: Research Development
and Clinical Applications.
Bradley, R. H., & Corwyn, R. F. (2008) Infant temperament, parenting, and external-
izing behavior in first grade: a test of the differential susceptibility hypothesis.
Journal of Child Psychology and Psychiatry, and Allied Disciplines, 49, 124-131.
Eggers, K., De Nil, L. F., & Van den Bergh, B. R. H. (2009) Factorial temperament
structure in stuttering, voice-disordered, and typically developing children.
Journal of Speech, Language, and Hearing Research, 52, 1610-1622. http://dx.doi.
org/10.1044%2F1092-4388%282009%2F07-0065%29.
Eggers, K., De Nil, L. F., & Van den Bergh, B. R. H. (2010) Temperament dimensions
in stuttering and typically developing children. Journal of Fluency Disorders, 35,
355-372. http://dx.doi.org/10.1016%2Fj.jfludis.2010.10.004.
Fisher, W. P. (1992) Reliability statistics. Rasch Measurement Transactions, 6, 238.
Gubbels, J. S., Kremers, S. P. J., Stafleu, A., Dagnelie, P. C., Goldbohm, R. A., De Vries,
N. K., & Thijs, C. (2009) Diet-related restrictive parenting practices: impact on
dietary intake of 2-year-old children and interactions with child characteristics.
Appetite, 52, 423-429. http://dx.doi.org/10.1016%2Fj.appet.2008.12.002.
Healey, D. M., Brodzinsky, L. K., Bernstein, M., Rabinovitz, B., & Halperin, J. M.
(2010) Moderating effects of neurocognitive abilities on the relationship be-
tween temperament and global functioning. Child Neuropsychology, 16, 20-31.
216 E. F. C. SLEdDens, et al.

Heesch, K. C., Mâsse, L. C., & Dunn, A. L. (2006) Using Rasch modeling to re-evaluate
three scales related to physical activity: enjoyment, perceived benefits and per-
ceived barriers. Health Education Research, 21, i58-i72.
Hughes, S. O., Shewchuk, R. M., Baskin, M. L., Nicklas, T. A., & Qu, H. (2008) Indul-
gent feeding style and children’s weight status in preschool. Journal of Develop-
mental and Behavioral Pediatrics, 29, 403-410. http://dx.doi.org/10.1097%2FDBP.0b
013e318182a976.
Kochanska, G., Aksan, N., & Joy, M. E. (2007) Children’s fearfulness as a moderator
of parenting in early socialization: two longitudinal studies. Developmental Psy-
chology, 43, 222-237.
Kochanska, G., De Vet, K., Goldman, M., Murray, K., & Putnam, S. P. (1994) Mater-
nal reports of conscience development and temperament in young children. Child
Development, 65, 852-868.
Kochanska, G., Murray, K., Jacques, T. Y., Koenig, A. L., & Vandegeest, K. A. (1996) In-
hibitory control in young children and its role in emerging internalization. Child
Development, 67, 490-507. http://dx.doi.org/10.1111%2Fj.1467-8624.1996.tb01747.x.
Majdandžić, M., & Van den Boom, D. C. (2007) Multimethod longitudinal assessment
of temperament in early childhood. Journal of Personality, 75, 121-168.
Majdandžić, M., Van den Boom, D. C., & Heesbeen, D. G. M. (2008) Peas in a pod:
biases in the measurement of sibling temperament? Developmental Psychology, 44,
1354-1368.
Miller, K. S., Vannatta, K., Compas, B. E., Vasey, M., McGoron, K. D., Salley, C. G., &
Gerhardt, C. A. (2009) The role of coping and temperament in the adjustment
of children with cancer. Journal of Pediatric Psychology, 34, 1135-1143. http://dx.doi.
org/10.1093%2Fjpepsy%2Fjsp037.
Nunnally, J. C., & Bernstein, I. H. (1994) Psychometric theory. (2nd ed.) New York:
McGraw-Hill.
Nygaard, E., Smith, L., & Torgersen, A. M. (2002) Temperament in children with
Down syndrome and in prematurely born children. Scandinavian Journal of Psy-
chology, 43, 61-71.
O’Connor, B. P. (2000) SPSS and SAS programs for determining the number of com-
ponents using parallel analysis and Velicer’s MAP test. Behavior Research Meth-
ods, Instrumentation, and Computers, 32, 396-402. http://dx.doi.org/10.1017%2
FS1368980009005916.
O’Connor, T. M., Hughes, S. O., Watson, K. B., Baranowski, T., Nicklas, T. A., Fisher,
J. O., Beltran, A., Baranowski, J. C., Qu, H., & Shewchuk, R. M. (2010) Parent-
ing practices are associated with fruit and vegetable consumption in pre-school
children. Public Health Nutrition, 13, 91-101.
Putnam, S. P., & Rothbart, M. K. (2006) Development of short and very short forms of
the Children’s Behavior Questionnaire. Journal of Personality Assessment, 87, 102-112.
Putnam, S. P., Sanson, A. V., & Rothbart, M. K. (2002) Child temperament and par-
enting. In M. H. Bornstein (Ed.), Handbook of parenting. Vol. 1. Mahwah, NJ: Erl-
baum. Pp. 255-277.
Rothbart, M. K., Ahadi, S. A., & Evans, D. E. (2000) Temperament and personality:
origins and outcomes. Journal of Personality and Social Psychology, 78, 122-135.
Rothbart, M. K., Ahadi, S. A., & Hershey, K. L. (1994) Temperament and social be-
havior in childhood. Merrill-Palmer Quarterly, 40, 21-39.
 Validation of CBQ and One-Item Temperament Scale 217

Rothbart, M. K., Ahadi, S. A., Hershey, K. L., & Fisher, P. (2001) Investigations of

temperament at three to seven years: the Children’s Behavior Questionnaire. Child
Development, 72, 1394-1408.
Rothbart, M. K., & Bates, J. E. (1998) Temperament. In W. Damon (Series Ed.) & N.
Eisenberg (Vol. Ed.), Handbook of child psychology: Vol. 3. Social, emotional and per-
sonality development. (5th ed.) New York: Wiley. Pp. 105-176.
Sha, M. M., & Pan, Y. (2009) The use of vignettes in evaluating multilingual ques-
tionnaires. Presented at the American Association for Public Opinion Research
Conference, May 14-17.
Sleddens, E. F. C., Kremers, S. P. J., Candel, M. J. J. M., De Vries, N. K., & Thijs, C.
(2011) Validating the Children’s Behavior Questionnaire in Dutch children: psy-
chometric properties and a cross-cultural comparison of factor structures. Psycho-
logical Assessment, 23, 417-426. http://dx.doi.org/10.1037%2Fa0022111.
Tumanova, V., Zebrowski, P. M., Throneburg, R. N., & Kulak Kayikci, M. E. (2011) Ar-
ticulation rate and its relationship to disfluency type, duration, and temperament
in preschool children who stutter. Journal of Communication Disorders, 44, 116-129.
Van den Bergh, B. R. H., & Ackx, M. (2003) Een Nederlandse versie van Rothbart’s
Children’s Behavior Questionnaire: interne consistentie en driefactoren model
van de subschalen [A Dutch version of Rothbart’s Children’s Behavior Question-
naire: internal consistency and three-factor model of the scales]. Kind en Adoles-
cent, 24, 77-84.
Van Zeijl, J., Mesman, J., Stolk, M. N., Alink, L. R., Van Ijzendoorn, M. H., Baker-
mans-Kranenburg, M. J., Juffer, F., & Koot, H. M. (2007) Differential suscepti-
bility to discipline: the moderating effect of child temperament on the association
between maternal discipline and early childhood externalizing problems. Journal
of Family Psychology, 21, 626-636.
Ward, T. M., Gay, C., Alkon, A., Anders, T. F., & Lee, K. A. (2008) Nocturnal sleep and
daytime nap behaviors in relation to salivary cortisol levels and temperament in
preschool-age children attending child care. Biological Research for Nursing, 9, 244-
253. http://dx.doi.org/10.1177%2F1099800407310158.
Wilson, M., Allen, D. D., & Li, J. C. (2006a) Improving measurement in health edu-
cation and health behavior research using item response modeling: comparison
with the classical test theory approach. Health Education Research, 21, i19-i32.
Wilson, M., Allen, D. D., & Li, J. C. (2006b) Improving measurement in health edu-
cation and health behavior research using item response modeling: introducing
item response modeling. Health Education Research, 21, i4-i18.
Wu, M., Adams, R., & Haldane, S. (2007) Conquest. Victoria, Australia: The Australian
Council for Educational Research.
Wu, T., Dixon, W. E., Dalton, W. T., Tudiver, F., & Liu, X. (2010) Joint effects of child
temperament and maternal sensitivity on the development of childhood obesity.
Maternal and Child Health Journal, 15, 469-477.
Zeller, M. H., Boles, R. E., & Reiter-Purtill, J. (2008) The additive and interactive
effects of parenting style and temperament in obese youth seeking treatment. In-
ternational Journal of Obesity, 32, 1474-1480.

Accepted January 9, 2012.