See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/262074431

Internal Structure of the Children Helping Out: Responsibilities,

Expectations, and Supports (CHORES) Measure

Article in American Journal of Occupational Therapy · May 2014

DOI: 10.5014/ajot.2014.010454 · Source: PubMed

CITATIONS READS

10 1,879

3 authors:

Louise Dunn Livia Magalhaes

Brenau University Federal University of Minas Gerais
22 PUBLICATIONS 514 CITATIONS 69 PUBLICATIONS 1,780 CITATIONS

SEE PROFILE SEE PROFILE

Marisa Cotta Mancini

Federal University of Minas Gerais
187 PUBLICATIONS 3,865 CITATIONS

SEE PROFILE

All content following this page was uploaded by Marisa Cotta Mancini on 11 November 2014.

The user has requested enhancement of the downloaded file.

Internal Structure of the Children Helping Out:
Responsibilities, Expectations, and Supports
(CHORES) Measure

Louise Dunn, Lı́via C. Magalhaes, Marisa Cotta Mancini

MeSH TERMS The purpose of this study was to examine the internal structure of the Children Helping Out: Responsibilities,
 activities of daily living Expectations, and Supports (CHORES), an assessment of household task participation for children. Rasch
analysis was used to examine patterns of item response and scale structure with data collected from care-
 adolescent
givers of 132 children and youth ages 6–14 yr with and without disabling conditions. Internal consistency
 child was strong for the total measure and the subscales. The items in both subscales fit the measurement model,
 housekeeping and the item difficulty order matched the expected pattern from harder to easier household task performance
 reproducibility of results and degree of caregiver assistance. The sample distribution in the hierarchical continuum showed that
 task performance and analysis younger participants and those with physical disabilities tended to score lower. Some inconsistencies in
rating scale use suggest a need for further clarification of the scoring criteria for measurement coherence.

Dunn, L., Magalhaes, L. C., & Mancini, M. C. (2014). Internal structure of the Children Helping Out: Responsibilities,
Expectations, and Supports (CHORES) measure. American Journal of Occupational Therapy, 68, 1–11. http://dx.
doi.org/10.5014/ajot.2014.010454

Louise Dunn, ScD, OTR/L, is Associate Professor,

Brenau University, Gainesville, GA 30501; ldunn@brenau.
edu. At the time of this study, she was Assistant Professor,
P articipation is a key predictor of quality of life and preparation for independent
living (Kanne et al., 2011; Law et al., 2006). Household task participation
warrants assessment because daily living activities are predictive of community
University of Utah, Salt Lake City.
living and greater independence for people of all ability levels (Kanne et al., 2011;
Lı́via C. Magalhaes, PhD, OTR, is Professor, Smith, Maenner, & Seltzer, 2012). Chores are common tasks that take place
Department of Occupational Therapy and Graduate frequently in all homes, and participation affords a way to examine skills and
Program in Rehabilitation Sciences, Universidade Federal
de Minas Gerais, Belo Horizonte, Brazil.
social participation (Goodnow, 1996; Paradise & Rogoff, 2009). In addition,
parents have reported that participation in domestic tasks is an important activity
Marisa Cotta Mancini, ScD, OT, is Professor, for their children (Adolfsson, Malmqvist, Pless, & Granuld, 2011).
Department of Occupational Therapy and Graduate
Engagement in household tasks helps prepare children for future living in the
Program in Rehabilitation Sciences, Universidade Federal
de Minas Gerais, Belo Horizonte, Brazil. community; such preparation occurs largely within the home (Anderson-Moore,
Chalk, Scarpa, & Vandivere, 2002; Paradise & Rogoff, 2009; Reid et al.,
2011). Thus, active involvement in domestic routines fosters children’s learning
of household tasks (Goodnow, 1996; Larson, 2004; Rogoff, 2003). The
learning may be tacit, such as by observing others perform household tasks, or
more explicit, such as with direct instruction and supervision.
Caregivers expect children to assume more responsibility for household tasks
as they age (Hofferth & Sandberg, 2001; Rogoff, 2003). Changes in re-
sponsibility are evident in the transfer of accountability for task completion
from the caregiver to the child (Rogoff, 2003). Documenting these changes
leading to children’s independence helps families adjust expectations and
identify strategies to involve and engage children in daily tasks. In addition,
exploration of children’s engagement in these tasks helps increase occupational
therapy practitioners’ understanding of how to prepare children of all abilities
for future independent or interdependent living.

The American Journal of Occupational Therapy 1

 Much variation occurs among families with regard to Preliminary analysis of the psychometric properties of
the household tasks that children perform because of the CHORES (Dunn, 2004) showed good test–retest
cultural values, the children’s ages, the children’s abilities, reliability (intraclass correlation coefficient [ICC] 5 .93)
and available time (Dunn, Coster, Orsmond, & Cohn, and strong internal consistency (r 5 .94). Occupational
2009; Goodnow, 2006; Hofferth & Sandberg, 2001; therapy practitioners and parents have supported the
Larson, 2004). Many caregivers value household task content validity of the CHORES (Dunn, 2004). The
participation as a way to help children learn to manage CHORES has differentiated performance and assistance
their future home; others (especially mothers) value the levels for children with cognitive and behavioral dis-
development of responsibility and the sense of being abilities such as attention deficit hyperactivity disorder
a part of the family (Anderson-Moore et al., 2002; Dunn (ADHD; Dunn et al., 2009), physical disabilities (Dunn
et al., 2009; Reid et al., 2011). & Gardner, 2013), and autism spectrum disorder (Dunn
Adaptive behavior measures, such as the Vineland & Gardner, 2013). Further examination of the internal
Adaptive Behavior Scale (VABS, Sparrow, Cicchetti, & structure and validity of the two subscales, Self-Care and
Balla, 2005), are considered the gold standard to assess Family Care, would strengthen the psychometric prop-
household task performance. The VABS is a normative erties of the CHORES. Thus, the purpose of this study
measure, however; it does not provide continuity in was to examine the internal structure and dimensions of
measuring the same tasks over time, nor does it provide the CHORES using Rasch analysis, including principal
a way to assess progress or change in assistance needed. A components analysis.
normative measure also may not be appropriate to docu-
ment changes in the performance of tasks that are Method
culturally laden. Families differ in expectations and satis-
faction and in their basis for distributing household Research Design
responsibilities, so a measure to target this construct A cross-sectional design was used with a convenience
needs to include tasks that are culturally sensitive and sample for this mailed survey study. Louise Dunn received
represent a range of complexity. Measurement of the ethical approval from internal review boards at Boston
tasks children perform and the amount of assistance they University and the University of Utah.
require or typically receive provides a way to examine
independence in home management. In addition, en- Participants
gagement in everyday tasks supports the social struc- The first author (Dunn) collected data from the caregivers
turing of the family environment (Anderson-Moore of 132 children and youth ages 6–14 yr with and without
et al., 2002; Dunn et al., 2009; Goodnow, 2006; disabilities from different regions of the United States.
Rogoff, 2003). The children formed three groups: (1) those developing
Assessment of responsibility for household tasks typically, (2) those with cognitive or behavioral disorders
provides a means to examine independence and its changes (ADHD, Asperger syndrome, or high-functioning au-
over time. Household tasks are routines that occur fre- tism), and (3) those with physical disabilities (cerebral
quently and are common in all homes; the learning of and palsy, spina bifida, orthopedic impairments, amputation).
active involvement in such tasks constitute a way to Caregivers were mailed a packet of questionnaires to
measure acquisition of independence among children complete on their own and a stamped, addressed return
and youth of various ability levels within and across envelope to return the questionnaires.
families. Dunn (2004) developed the Children Helping Participants were recruited through multiple strate-
Out: Responsibilities, Expectations, and Support gies, including ads posted on email lists, at recreation
(CHORES) measure to examine participation in the centers, and in specialty pediatric clinics and through
home in common household tasks. The CHORES as- colleagues and occupational therapy students. To be in-
sesses domestic task engagement with two subscales: Self- cluded in the study, the targeted child had to have
Care and Family Care. Self-Care subscale items examine caregiver-reported intellect in the normal range, and the
involvement of the child or youth in organizing and family had to include between 2 and 4 children. Exclusion
caring for his or her belongings in his or her own space, criteria included caregiver report of the child’s intellectual
whereas Family Care subscale items examine the in- disability, seizures, sensory impairments (e.g., blind,
volvement of the child or youth in looking after the needs deaf), or inability to use the upper extremities and pres-
and belongings of family members in common spaces in ence of any sibling who was not able to participate in
the home. activities with same-age peers.

2 May/June 2014, Volume 68, Number 3

 During recruitment, the first author sought partic- of intellectual functioning. Participants completed the
ipants who had children in the three groups. Group questionnaires and mailed them to the first author in the
members were matched by child’s age, gender, and eth- stamped, addressed envelopes provided. Participants re-
nicity. The sample was further divided into two age ceived a gift certificate in compensation for their time and
groups—younger children ages 6–10 yr and older chil- assistance. The CHORES subscales were used in multiple
dren ages 11–14 yr. studies, and the data were combined to conduct the
Rasch analyses of the instrument.
Instrument
The CHORES is a 34-item questionnaire designed to Data Analysis
assess participation in household tasks (Performance scale) To investigate the measurement qualities of the CHORES,
and help needed to do household tasks (Assistance scale). two different approaches were used: (1) Cronbach’s a was
The Performance scale has a dichotomous yes–no response used to test the internal consistency of the subscales, and
format. Total performance is calculated by summing all (2) the Rasch model was applied to conduct a more de-
the yes responses. Higher performance scores indicate tailed item analysis to examine the structure of the
greater extent of age-expected participation. The response CHORES subscales. The Rasch model derives from item
format of the Assistance scale is a 7-point Likert scale response theory and has been used in the construction
with 0 5 child is not expected to perform task, 1 5 child and validation of several rehabilitation- and health-
cannot perform task, 2 5 child performs task with a lot of related measures (Bedell, 2009; Haley, Coster, Ludlow,
help, 3 5 child performs task with some help, 4 5 child Haltiwanger, & Andrellos, 1992; Russell et al., 2000).
performs task with supervision, 5 5 child performs task Rasch analysis is based on a probabilistic model that
given verbal cue, and 6 5 child performs task on own converts ordinal scores obtained from rating scales into
initiative more than 50% of the time. CHORES Assistance interval-level measures. The model is based on the prin-
scores are calculated by summing all applicable item re- ciple that only two attributes are needed to determine
sponses and dividing this number by the total possible participants’ response to test items: the ability of the
score from all applicable items; the result is multiplied by participant and the difficulty of the test item, each ex-
100 to conform to a 100-point scale. Higher Assistance pressed as estimates of the underlying latent trait. How-
scores indicate greater independence with household ever, this principle holds only if the items measure a
tasks. unidimensional construct.
In addition, the CHORES has two subscales: Self- Based on these principles, it is possible to construct
Care household tasks, in which item content primarily a measure, like a ruler, that is defined by items hierarchically
affects the child (e.g., Cleans up after own play), and organized from easy to difficult. Because participants and
Family Care household tasks, in which item content af- items are calibrated along the same interval metric, expressed
fects other members of the family (e.g., Prepares a cold in logits and distributed in hierarchical order, it is possible
meal for family). These subscales illustrate the conceptu- to examine the extent to which items are of appropriate
alization that household participation includes family difficulty for the sample. The logic of the distribution of
division of labor and that members’ take responsibility on indicated item and person hierarchies can be used as evi-
a routine basis for tasks related to self-maintenance and dence of construct validity (Chien & Bond, 2009).
family maintenance (Gager, Sanchez, & Demaris, 2009; Rasch analysis using the Winsteps program (Version
Goodnow, 1988; Hofferth & Sandberg, 2001). The 3.72.0; Linacre, 2013) was used to examine whether
CHORES has strong stability (test–retest, ICC 5 .94) CHORES items combine to measure a single unidi-
and internal consistency (r 5 .93; Dunn, 2004). Parents mensional construct. Analysis using the rating scale
and occupational therapists had 80% agreement on model indicated that increments in percent of response in
the CHORES items, supporting its content validity each score category of the CHORES rating scale varied,
(Dunn, n.d.). suggesting irregularity in the difficulty intervals across
items. Therefore, we proceeded with a partial-credit
Procedure model. CHORES subscales’ unidimensionality was ex-
The first author contacted the participants, obtained amined using two methods: (1) principal components
consent, and conducted a brief screening. During the analysis, with variance explained by the first dimension
screening, she asked participants about their child’s >50% and eigenvalue of the unexplained variance of the
classroom placement, communication abilities, and ad- first residual factor <2, and (2) goodness of fit of items
ditional concerns (e.g., seizures) to support their reports and participants to the Rasch model, as expressed by the

The American Journal of Occupational Therapy 3

residuals derived from the interval-level scaling (<5% of The structure of the Assistance 7-point scoring system
both the infit and outfit mean square [MnSq] values were was also investigated to verify whether the scoring criteria
³1.4 with associated t ³ ±2.0). were behaving according to the principles of the Rasch
Results from a preliminary principal components model. The 0 scores were excluded because these scores
analysis helped guide further data analysis. The CHORES indicate that the caregivers did not expect their children do
items fell into at least two groups or dimensions that the tasks. The response pattern of each scoring category for
resembled the natural division of the measure into Self- each item was examined to identify weaknesses and points
Care and Family Care subscales (Dunn, 2004). Because that needed to be revised or clarified.
only a couple of items had been misplaced by the load-
ings, we decided to keep the two original dimensions Results
(subscales) of the CHORES because these dimensions
made theoretical sense. Results indicated that although Participants
variance explained for each scale was >50%, the ei- Caregivers of 132 children of all abilities ages 6–14 yr
genvalues were >2.0 (i.e., 2.2 for the Self-Care and 2.9 for participated in this study. Of the children, 49.2% (n 5
the Family Care subscale), implying extra dimensions; 65) were in the typically developing group, 33.3% (n 5
however, these extra dimensions did not make either 44) were in the cognitive or behavioral disorders group,
theoretical or clinical sense, and when analyzed sepa- and 17.4% (n 5 23) were in the physical disabilities
rately, the reliability values fell below .80. Thus, we de- group. Approximately 60% (n 5 79) of the children were
cided to analyze the full subscales to identify misfitting in the younger age group. The numbers of male (49%,
items that were not contributing to defining the measure. n 5 64) and female children were approximately equal.
Items that fit the model are those that present a pattern Table 1 provides demographic information on the sample
of response that conforms to the expectations of the hi- and each group’s mean scores on the CHORES.
erarchical model: Children with higher ability should do
more chores and, therefore, score higher on most items, Internal Consistency
easy or hard, and children with lower ability should score
lower on the difficult items. The Winsteps program Both the Self-Care and Family Care subscales showed
provides fit statistics that indicate the fitness of the pattern evidence of strong internal consistency (as 5 .96 and .98,
of responses to the model. Items with infit and outfit respectively).
statistics ³1.4 (MnSq) and with associated t ³ ±2.0
(Z standardized distribution value) are considered to CHORES Self-Care Subscale
misfit and should be flagged for revision and further Rasch Analysis—Scalability and Dimensionality. Re-
analysis. If no more than 5% of the items meet the above garding calibration, all 13 items from the CHORES
criteria for misfitting, the subscale is considered valid. Self-Care subscale had infit and outfit statistics within
The infit statistics and children pattern of responses to acceptable Rasch measurement ranges, and only 6 children
the model, along with analysis of the logic of the hierarchy (4.5%) misfitted the model. Item and person reliability
of item difficulty levels, were used as an indicator of were .88 and .95, respectively, and the items separated
construct validity. Inspection of the value of the mean the children into 2.45 ability levels. Table 2 presents
person measure and of the item–person maps was con- item calibration and fit statistics for the CHORES Self-
ducted as a means to examine the relationships among Care subscale.
item difficulty, the ability of the children (item targeting), Rasch Analysis—Hierarchical Ordering and Test Targeting.
and the spread of the items over the range of ability levels. Item difficulty calibration spanned 1.37 logits. The easiest
The Winsteps program also calculates the reliability was Item 17, Puts own laundry in hamper (logit 5 2.71),
for item calibration and corresponding person measure, and the most difficult was Item 9, Makes self a hot meal
with values above .80 considered an indication of the (logit 5 .66). The mean measure fell 1.09 logits above the
stability of the measures. The error for each child measure mean item calibration, indicating that, on average, the
and item calibration was calculated. The reliability and CHORES tasks were easy relative to the children’s ability
error values are used to calculate the separation index, or levels. The number of respondents (n) in Table 2 indicates
number of ability levels the items divide the sample into, the relevance of each task to the parents; items considered
which should be at least two levels (Bond & Fox, 2007). irrelevant (score 5 0) by specific parents were dropped
Cronbach’s a was set at >.80 to indicate the subscales’ from the analysis. Although most parents (98.5%, n 5
internal consistency. 130) considered Puts own laundry in hamper (Item 17) to

4 May/June 2014, Volume 68, Number 3

Table 1. Demographic and Clinical Characteristics of Caregivers and Children and CHORES Scores
Characteristics n %
Child’s age category
Younger (6–10 yr) 79 59.8
Older (11–14 yr) 53 40.2
Child’s sex
Male 64 48.5
Female 68 51.5
Child’s race or ethnicity
White 124 94.0
Hispanic 8 6.0
Mother’s education
High school diploma 10 7.6
Some college 47 35.6
Undergraduate degree 48 36.4
Graduate education 27 20.5
Disability group
Cognitive or behavioral disorders 44 33.3
Physical disabilities 23 17.4
Typically developing 65 49.2
Parent-reported child intellect
Low average 5 3.8
Average 61 46.2
Above average 66 50.0

CHORES Scores
Performance Assistance
Age Category and Disability Group n (%) M (SD) POMP M (SD)
Younger, cognitive or behavioral disorders 26 (19.7) 20.88 (4.52) 70.89 (12.04)
Older, cognitive or behavioral disorders 18 (13.6) 25.22 (4.82) 71.73 (9.44)
Younger, physical disabilities 13 (9.8) 19.31 (6.43) 59.48 (19.07)
Older, physical disabilities 10 (7.6) 25.00 (6.13) 74.53 (16.42)
Younger, typically developing 40 (30.3) 22.58 (4.82) 77.99 (11.24)
Older, typically developing 25 (18.9) 29.16 (4.17) 88.49 (5.14)
Note. CHORES 5 Children Helping Out: Responsibilities, Expectations, and Supports; M 5 mean; POMP 5 percent of maximum possible; SD 5 standard deviation.

be relevant, fewer (60.6%, n 5 80) considered Item 23, (MnSQ < 1.4), and 17 of 21 items had acceptable outfit
Dusts own room, to be relevant. Considering the scoring statistics values (MnSQ < 1.4); however, no item pre-
criteria, most children performed the Self-Care items with sented values for both infit and outfit outside the criteria
a verbal cue (score of 5; 33.3%, n 5 44) or on their own for misfitting items. Only four children (3.0%) misfit the
initiative (score of 6; 37.1%, n 5 49); very few were model. Item and person reliability were .92 and .87, re-
considered unable to do specific tasks (score of 1; 3.0%, spectively, and the items separated the children into 2.57
n 5 4). levels of ability.
Figure 1 shows the task difficulty hierarchy, on the Rasch Analysis—Hierarchical Ordering and Test
right, and the distribution of children according to ability Targeting. Item difficulty calibration spanned 2.06
level on the Self-Care subscale, on the left. The children logits and ranged from 21.01 (easiest: Item 27, Feeds pet)
are organized in a normal-like distribution, dislocated to 20.99 (most difficult: Item 21, Runs washer/dryer).
somewhat upward (1.09 logits) in relation to the item The mean measure was 0.54 logits above the mean item
distribution. calibration, and as shown in Figure 2, the Family Care
items were more targeted to the distribution of the
children’s ability to do family care tasks than the Self-
CHORES Family Care Subscale Care items. However, more Family Care items were
Rasch Analysis—Scalability and Dimensionality. As seen considered beyond expectations; for example, Item 20,
in Table 2, all items on the CHORES Family Care Puts laundry away for family, was scored by only 27.3%
subscale had infit statistics within acceptable limits (n 5 36) of the parents. The scoring system showed the

The American Journal of Occupational Therapy 5

Table 2. Item Calibration and Fit Statistics for the Self-Care and Family Care Subscales of the CHORES
Infit Outfit
a
Item n (%) Item Difficulty (logit) SE Mnsq Z Mnsq Z
Self-Care subscale
9. Makes self a hot meal 86 (63.7) 0.66 0.10 1.26 1.6 1.21 1.1
22. Sweeps or vacuums own room 102 (75.6) 0.56 0.10 1.02 0.2 0.95 20.2
31. Organizes after-school belongings 123 (91.1) 0.25 0.08 0.82 21.4 0.71 21.5
23. Dusts own room 82 (60.7) 0.19 0.13 1.28 1.4 1.38 1.7
2. Picks up own bedroom 133 (98.5) 0.13 0.09 0.87 21.0 0.84 21.2
1. Cleans up after own play 134 (99.3) 0.12 0.11 0.95 20.3 0.87 21.0
5. Puts away own clothes 125 (92.3) 0.07 0.10 1.03 0.2 1.03 0.2
30. Organizes own belongings for school 128 (94.8) 0.03 0.08 1.04 0.4 0.99 0.0
3. Makes own bed 121 (89.6) 20.05 0.09 1.28 1.7 1.14 0.8
18. Puts away own clean laundry 115 (85.2) 20.14 0.12 0.76 21.4 0.78 21.3
7. Makes self a cold meal 121 (89.6) 20.45 0.10 1.09 0.6 1.48 1.5
6. Makes self a snack 130 (96.3) 20.68 0.10 0.89 20.6 1.49 1.2
17. Puts own laundry in hamper 133 (98.5) 20.71 0.12 0.85 20.9 0.70 21.4
Family Care subscale
21. Runs washer/dryer 54 (40.0) 1.05 0.14 0.92 20.3 0.86 20.3
10. Prepares part of a hot meal for family 78 (57.8) 0.51 0.12 0.92 20.5 1.02 0.2
20. Puts laundry away for family 36 (27.3) 0.50 0.21 0.70 21.0 0.61 21.1
33. Runs errand 83 (61.5) 0.45 0.10 0.60 22.2 0.75 20.8
4. Picks up area shared by others 130 (96.3) 0.25 0.11 0.98 20.1 0.94 20.3
19. Sorts laundry for family 60 (44.4) 0.20 0.15 1.21 1.1 1.20 0.9
28. Cares for younger sibling 73 (54.1) 0.18 0.11 0.94 20.3 0.94 20.2
15. Takes out garbage/recycling 111 (82.2) 0.17 0.10 1.05 0.4 1.81 2.9
16. Cleans bathroom 73 (54.1) 0.13 0.12 0.97 20.1 1.32 1.7
29. Cares for other family members 59 (43.7) 20.02 0.13 1.24 1.1 1.26 1.0
24. Sweeps or vacuums home 100 (74.1) 20.02 0.11 0.94 20.3 0.86 20.7
13. Washes dishes (loads dishwasher) 101 (74.8) 20.04 0.11 0.99 0.0 0.84 20.9
14. Dries dishes (unloads dishwasher) 94 (69.6) 20.06 0.12 0.85 20.9 0.71 21.3
26. Cares for plants 57 (42.2) 20.11 0.17 1.24 1.1 1.92 2.9
12. Brings in or puts away groceries 117 (86.7) 20.11 0.11 0.89 20.6 0.72 21.4
32. Takes a phone message 98 (72.6) 20.21 0.09 1.33 1.8 1.48 1.7
25. Dusts the house 67 (49.6) 20.30 0.14 1.13 0.7 1.09 0.5
8. Prepares a cold meal for family 93 (68.9) 20.44 0.12 0.87 20.9 0.80 21.4
34. Gets the mail or the newspaper 116 (85.9) 20.50 0.11 1.32 1.4 1.55 2.2
11. Sets or clears the table 125 (92.6) 20.62 0.13 0.80 21.1 0.82 21.1
27. Feeds pet 103 (76.3) 21.01 0.14 0.86 20.6 0.90 20.5
Note. CHORES 5 Children Helping Out: Responsibilities, Expectations, and Supports; Mnsq 5 mean square; SE 5 standard error; Z 5 standardized distribution
value.
a
Number (percentage) of children (N 5 132) that were scored on the item; scores of 0 (child is not expected to perform task) were deleted from the analysis.

same characteristics as the Self-Care subscale, with the the internal structure of the CHORES. The CHORES
score of 1 (Child cannot perform task) being used very is a unique tool; it is one of the few measures that
rarely (6.1%, n 5 8) and the score of 5 (Child performs examine the participation of children in common home
task given verbal cue) being the most frequent score activities, including social engagement as a family
(50.8%, n 5 67). member, which may be predictive of future independent
living.
Findings from the Rasch analysis substantiate the
Discussion unidimensionality of the Self-Care and Family-Care
This study expands on the earlier examination of test subscales thus supporting the use of summed subscale
stability, internal consistency, and content validity of scores rather than a total score when measuring children’s
the CHORES, a measure of household task participa- assistance with household tasks. In addition, these find-
tion for children of all abilities ages 6–14 yr (Dunn, ings support the internal consistency of the CHORES’
2004). In this study, we used Rasch analysis to examine Assistance ratings, with a larger sample.

6 May/June 2014, Volume 68, Number 3

Figure 1. Map of CHORES Self-Care item difficulty and children ability hierarchy.

The American Journal of Occupational Therapy 7

Figure 2. Map of CHORES Family Care item difficulty and children ability hierarchy.

8 May/June 2014, Volume 68, Number 3

 The sample distribution in the hierarchical contin- questionnaire. Clarification of each scoring criteria and
uum suggested a good fit of children by ability and age; rating, with a brief explanatory sentence may help parents
younger participants and those with physical disabilities make more precise choices, especially considering that
tended to score lower. However, the distribution of items this instrument could be completed online. For example,
and children along the hierarchical continuum was distinct performing task with a verbal prompt (rating 5) and
for the two subscales. Although the Self-Care subscale performing task with supervision (rating 4) might be
presented a ceiling effect, with 2 children from the older clearer if presence of caregiver during the task was de-
group receiving maximum scores, items from the Family fined. Here, presence of family member throughout the
Care subscale were more difficult, with maximum item child’s performance of the task would describe with su-
difficulty exceeding the ability level of the participants pervision and performance of the task without family
(Figures 1 and 2). The values of the mean measure for the member would be the rating for the verbal prompt.
children in each subscale were consistent with this pat- Bourke-Taylor, Law, Howie, and Pallant (2009) pro-
tern; children scored 1.09 logits above the mean item vided assistance ratings for leisure activities that would be
calibration on the Self-Care subscale and 0.54 logits on a helpful resource for refining the rating criteria for the
the Family Care subscale, indicating that the Self-Care CHORES.
subscale items, on average, were easier for the sample. Comparing the response pattern on the two scales,
Although the Rasch model is structured to deal with 51% percent of the scores on the Family-Care subscale
missing data in the analysis, it is noteworthy that parents were the given verbal cue rating, whereas only 33% of the
scored more tasks on the Family Care subscale as 2 (Child Self-Care subscale items received this score. Such a dis-
is not expected to perform task) than on the Self-Care tinction on the same rating across subscales evidences the
subscale. As shown in Table 2, more parents responded to greater complexity of skills needed to initiate and com-
the Self-Care items (ranging from 63.7% to 99.3%) than plete Family Care tasks. In addition, caregivers reported
to the Family Care items (ranging from 27.3% to that they did not expect their children to do the Family
96.3%). These results reflect what we have observed Care tasks more often than Self-Care tasks, suggesting
clinically: Most parents require children to do self-care that families of this cultural background may value chores
chores, but assignment of family care chores varies, is on these two subscales distinctly. The extent to which
more flexible, and depends on cultural values and ex- culture influences children and adolescent engagement in
pectations (Goodnow, 1996, 2006). self-care and family care tasks remains to be investigated.
Hierarchies of tasks fit the expected pattern of Parents of children of all abilities have reported that
a continuum of difficulty for household tasks. Most of the participation in home activities, including household
parents reported that their children performed easier tasks, tasks, was important to them (Adolfsson et al., 2011; Reid
such as Cleans up after own play, many independently or et al., 2011). Parents of children with special needs have
with a verbal cue. Fewer children performed the more reported that participation in household tasks was im-
difficult tasks, such as Prepares part of a hot meal for portant but often was not addressed by current service
family, and often with assistance. The availability of delivery models (Bedell, Khetani, Cousins, Coster, &
a difficulty hierarchy for household tasks is helpful to Law, 2011; Law et al., 2013). Household task partici-
guide clinical reasoning and to inform caregivers about pation warrants attention from occupational therapy
the degree of involvement in household tasks to expect of practitioners because parents perceive that it contributes
children. to their family routines, is a way to help their children
The rating scale presents some concerns with respect develop responsibility, and prepares their children for the
to a contracted degree of difficulty for both CHORES future (Dunn et al., 2009; Dunn & Gardner, 2013; Reid
Self-Care subscale and Family-care subscale items. Spe- et al., 2011).
cifically, the rating scale includes a combination of two These concerns support the utility of a valid and
criteria—that is, two ratings to identify why the child reliable measure of household task participation such as
does not do a specific task and five ratings to identify the the CHORES. The psychometric properties of the
type and amount of assistance frequently made available CHORES are adequate to make it useful in measuring
in the family daily routine. To examine the unidimen- children’s participation in household tasks. Incorporation
sionality of the CHORES, these criteria were combined of the CHORES into the assessment protocol offers
into one rating scale for the scoring of each item. Al- a springboard to promote discussion with caregivers and
though the presence of two criteria enriches the items their children about engaging in family routines and
scoring, it may pose confusion to parents who fill out the preparing for independent living. Addressing concerns by

The American Journal of Occupational Therapy 9

analyzing task difficulty and environmental supports and References
barriers to participation can inform practice and help Adolfsson, M., Malmqvist, J., Pless, M., & Granuld, M.
practitioners and families prepare children of all abilities (2011). Identifying child functioning from an ICF–CY
for future independent living. perspective: Everyday life situations explored in measures
of participation. Disability and Rehabilitation, 33, 1230–
1244. http://dx.doi.org/10.3109/09638288.2010.526163
Limitations and Future Research Anderson-Moore, K., Chalk, R., Scarpa, J., & Vandivere, S.
The study participants were from a relatively homo- (2002). Family strengths: Often overlooked, but real (Re-
geneous population, mostly White and from the United search brief). Washington, DC: Child Trends. Retrieved
States. Initial data collection did not include repre- from http://www.childtrends.org./?publications5family-
strengths-often-overlooked-but-real
sentation of children with intellectual disabilities. We
Bedell, G. (2009). Further validation of the Child and Adoles-
used a mail survey to gather information and interviews cent Scale of Participation (CASP). Developmental Neuro-
with some participants to confirm their understanding rehabilitation, 12, 342–351. http://dx.doi.org/10.3109/
of the ratings. Development of an electronic version of 17518420903087277
the CHORES warrants consideration, given the ease of Bedell, G. M., Khetani, M. A., Cousins, M. A., Coster, W. J.,
access and completion that technology affords care- & Law, M. C. (2011). Parent perspectives to inform de-
velopment of measures of children’s participation and
givers. A CHORES website could also offer opportu-
environment. Archives of Physical Medicine and Rehabilita-
nities for providing links to resources and a broader tion, 92, 765–773. http://dx.doi.org/10.1016/j.apmr.2010.
discussion network for supporting and engaging chil- 12.029
dren of all abilities in household tasks. Studies with Bond, T. G., & Fox, C. M. (2007). Applying the Rasch model:
larger numbers of children of varying abilities would Fundamental measurement in the human sciences (2nd ed.).
further support the construct and discriminative validity Mahwah, NJ: Erlbaum.
of the CHORES. Bourke-Taylor, H., Law, M., Howie, L., & Pallant, J. F.
(2009). Development of the Assistance to Participate Scale
(APS) for children’s play and leisure activities. Child: Care,
Implications for Occupational Health and Development, 35, 738–745. http://dx.doi.org/10.
Therapy Practice 1111/j.1365-2214.2009.00995.x
Chien, C. W., & Bond, T. G. (2009). Measurement properties
The findings of this study have the following implications of fine motor scale of Peabody Developmental Motor
for occupational therapy practice: Scales–second edition: A Rasch analysis. American Journal
• The CHORES provides practitioners with a tool that of Physical Medicine and Rehabilitation, 88, 376–386.
is psychometrically sound, easy to administer, rela- http://dx.doi.org/10.1097/PHM.0b013e318198a7c9
Dunn, L. (n.d.). [Data collected from one focus group of 7 oc-
tively brief, and clinically relevant to measure chil-
cupational therapists and two focus groups of 5 to 6 parents
dren’s participation in and degree of assistance needed each during 2002 and 2003]. Unpublished raw data.
to participate in household tasks. Dunn, L. (2004). Validation of the CHORES: A measure of
• The CHORES measures a valued aspect of home school-aged children’s participation in household tasks.
participation that contributes to children’s roles in Scandinavian Journal of Occupational Therapy, 11,
the family and to their preparation for independent 179–190. http://dx.doi.org/10.1080/11038120410003673
living. Dunn, L., Coster, W. J., Orsmond, G. I., & Cohn, E. S.
(2009). Household task participation of children with
• The CHORES provides a way to measure some as-
and without attentional problems. Physical and Occupa-
pects of preparation for adult roles for children. tional Therapy in Pediatrics, 29, 258–273. http://dx.doi.
• The CHORES provides information about relevant org/10.1080.01942630903008350
household tasks with a range of complexity. Dunn, L., & Gardner, J. (2013). Household task participation
• The internal structure of the CHORES discriminates of children with and without physical disability. American
household task participation among children by ability Journal of Occupational Therapy, 67, e100–e105. http://dx.
and age. s doi.org/10.5014/ajot.2013.008102
Gager, C., Sanchez, L., & Demaris, A. (2009). Whose time
is it? The effect of employment and work/family
Acknowledgments stress on children’s housework. Journal of Family
Issues, 30, 1459–1485. http://dx.doi.org/10.1177/
We appreciate the participation and contributions of the
0192513X09336647
caregivers to our study. We also appreciate the review by Goodnow, J. J. (1988). Children’s household work: Its nature
Tina McNulty and recruitment assistance from practi- and functions. Psychological Bulletin, 103, 5–26. http://dx.
tioners, community agencies, and graduate students. doi.org/10.1037/0033-2909.103.1.5

10 May/June 2014, Volume 68, Number 3

Goodnow, J. J. (1996). From household practices to parents’ Law, M., King, G., King, S., Kertoy, M., Hurley, P., Rosenbaum,
ideas about work and interpersonal relationships. In S. P., . . . Hanna, S. (2006). Patterns of participation in
Harkness & C. M. Super (Eds.), Parents’ cultural belief recreational and leisure activities among children with
systems: Their origins, expressions, and consequences (pp. complex physical disabilities. Developmental Medicine
313–344). New York: Guilford Press. and Child Neurology, 48, 337–342. http://dx.doi.org/10.
Goodnow, J. J. (2006). Cultural perspectives and parents’ 1017/S0012162206000740
views of parenting and development: Research directions. Linacre, M. J. (2013). Winsteps: Rasch-model computer program.
In K. H. Rubin & O. Chung (Eds.), Parenting beliefs, Chicago: MESA Press.
behaviors, and parent–child relations: A cross-cultural per- Paradise, R., & Rogoff, B. (2009). Side by side: Learning by
spective (pp. 35–57). New York: Psychology Press. observing and pitching in. Ethos (Berkeley, Calif.), 37,
Haley, S. M., Coster, W. J., Ludlow, L. H., Haltiwanger, J. T., 102–138.
& Andrellos, P. J. (1992). Pediatric Evaluation of Disabil- Reid, A., Imrie, H., Brouwer, E., Clutton, S., Evans, J.,
ity Inventory (PEDI). Boston: New England Medical Cen- Russell, D., & Bartlett, D. (2011). “If I knew then what
ter Hospitals. I know now”: Parents’ reflections on raising a child with
Hofferth, S. L., & Sandberg, J. F. (2001). How American cerebral palsy. Physical and Occupational Therapy in Pediat-
children spend their time. Journal of Marriage and the rics, 31, 169–183. http://dx.doi.org/10.3109/01942638.
Family, 63, 295–308. http://dx.doi.org/10.1111/j.1741- 2010.540311
3737.2001.00295.x Rogoff, B. (2003). Cultural nature of human development. New
Kanne, S. M., Gerber, A. J., Quirmbach, L. M., Sparrow, S. S., York: Oxford University Press.
Cicchetti, D. V., & Saulnier, C. A. (2011). The role of Russell, D. J., Avery, L. M., Rosenbaum, P. L., Raina, P. S.,
adaptive behavior in autism spectrum disorders: Implica- Walter, S. D., & Palisano, R. J. (2000). Improved scaling
tions for functional outcome. Journal of Autism and De- of the Gross Motor Function Measure for children with
velopmental Disorders, 41, 1007–1018. http://dx.doi.org/10. cerebral palsy: Evidence of reliability and validity. Physical
1007/s10803-010-1126-4 Therapy, 80, 873–885.
Larson, E. A. (2004). Children’s work: The less-considered Smith, L. E., Maenner, M. J., & Seltzer, M. M. (2012).
childhood occupation. American Journal of Occupational Developmental trajectories in adolescents and adults
Therapy, 58, 369–379. http://dx.doi.org/10.5014/ajot.58. with autism: The case of daily living skills. Journal
4.369 of the American Academy of Child and Adolescent Psy-
Law, M., Anaby, D., Teplicky, R., Khetani, M., Coster, W., & chiatry, 51, 622–631. http://dx.doi.org/10.1016/j.jaac.
Bedell, G. (2013). Participation in the home environment 2012.03.001
among children and youth with and without disabilities. Sparrow, S., Cicchetti, D., & Balla, D. (2005). Vineland Adap-
British Journal of Occupational Therapy, 76, 58–66. http:// tive Behavior Scales II. Circle Pines, MN: American Guid-
dx.doi.org/10.4276/030802213X13603244419112 ance Services.

The American Journal of Occupational Therapy 11

 AUTHOR QUERIES

AUTHOR PLEASE ANSWER ALL QUERIES

• Note that these are uncorrected page proofs that do not reflect proofreader corrections or corrections
submitted after the article went to typesetting. If you submitted corrections and do not see them
reflected in the document, please indicate what needs to be corrected.
• On p. 4, top of col. 1, you note: outfit mean square [MnSq] values were $ 1.4 with associated t $
62.0. Should the 6 be deleted? The same expression appears in the second full paragraph in this
column.
• On Figure 1, at the top, what does P stand for? Should the text read “children9s ability level” rather
than “children ability level,” here and in Figure 2?
For Figures 1 and 2, ok to create a separate legend and move the following into the caption:
“Representation of the children ability levels, on the left, and item calibration, on the right. Symbols
represent children from study sample. Arrows indicate gaps on children ability level not targeted by test
items.”

View publication stats