SCIENTIFIC/CLINICAL ARTICLES

Measuring Dexterity in Children Using the

Nine-hole Peg Test

Janet L. Poole, PhD, OTR/L ABSTRACT: The purpose of this study was to measure dexterity
in children aged 4–19 years using the Nine-hole Peg Test. Four
Patricia A. Burtner, PhD, OTR/L hundred and six children were tested with their dominant hand
Theresa A. Torres and then their nondominant hand. A commercial version of the
Occupational Therapy Graduate Program Nine-hole Peg Test was used. An analysis of variance showed a
Department of Orthopaedics and Rehabilitation main effect for age, gender, and hand dominance. Speed of
dexterity improved with age. In all age groups, females performed
University of New Mexico faster than males. Participants performed faster with the dominant
Albuquerque, New Mexico hand than the nondominant hand. The normative data collected
provide information for comparing scores to children with differ-
ent diagnostic categories to screen for fine motor difficulties.
Cheryl Kirk McMullen, OTR/L J HAND THER. 2005;18:348–351.
EASI
Albuquerque, New Mexico

Amy Markham, OTR/L

Albuquerque Public Schools
Albuquerque, New Mexico

Michelle Lee Marcum, OTR/L

Ridgecrest Healthcare
Phoenix, Arizona

Jennifer Bradley Anderson, OTR/L

Education Assessment Systems Inc.
Shiprock, New Mexico

Clifford Qualls, PhD

Clinical Research Center
University of New Mexico School of Medicine
Albuquerque, New Mexico

The performance of many tasks of daily living, for screening and proposed that the Nine-hole Peg
school activities, and play require fine motor dexter- Test (9-HPT) may be an appropriate screening tool.
ity. Although it has been reported that 10% of school- The 9-HPT is a timed test in which nine pegs are
age children have difficulty with fine motor tasks,1 inserted and removed from nine holes in the peg-
fine motor screening is not routinely performed for board with each hand. The original norms for the
school age children. Smith et al.2 suggested that one 9-HPT for adults were published by Kellor et al.3 At
explanation may be the lack of a simple and easy tool that time, the pegboard was not commercially avail-
able and required construction of the pegboard and
pegs from wood. Since that time, several commercial
Portions of this manuscript were presented at the New Mexico
Occupational Therapy Association Conference in Albuquerque, versions of the test have been marketed.4
NM, November 2001. The Smith et al.2 study developed normative data
Supported in part by the Clinical Research Center at UNM School for children 5–10 years of age using a modified
of Medicine (Grant M01RR0097). version of the Sammons, Preston and Rolyan peg
Address correspondence and reprint request to Dr. Poole, test. In this study, high interrater reliability (rs > 0.99)
Occupational Therapy Graduate Program, Department of
and test–retest reliability (rs = 0.81 and 0.79) were
Orthopaedics and Rehabilitation, University of New Mexico,
MSC09 5240, Albuquerque, NM 87131-0001; e-mail: <jpoole@ reported. Furthermore, concurrent validity of the
salud.unm.edu>. 9-HPT was examined by correlating scores on the
doi:10.1197/j.jht.2005.04.003 9-HPT with scores on the Purdue Pegboard, which

348 JOURNAL OF HAND THERAPY

yielded a correlation coefficient of 0.80 for the dom- TABLE 1. Subject Characteristics: Age, Sex, and Hand
inant hand and 0.74 for the nondominant hand.2 The Dominance
mean completion time was also shown to be different Males Females
for typical developing children and children in spe- Mean
cial education.2 Thus, the 9-HPT has the potential to Age age Mean
be a quick and easy to administer tool for screening (yr) N (yr) Right* Left* N age Right* Left*
fine motor problems in children. However, these 4–5 27 5.0 24 3 21 5.0 17 4
researchers used a modified version of the commer- 6–7 25 6.8 21 4 23 7.1 20 3
cially available pegboard for the normative, reliabil- 8–9 23 9.1 19 4 26 8.9 23 3
ity and validity study. Thus, the scores in the Smith 10–11 24 11.0 24 0 21 11.0 18 3
12–13 23 12.9 23 2 24 12.9 23 1
et al. study may not be representative of scores on the
14–15 25 15.1 21 4 25 14.9 24 1
commercially available version.4 Another study did 16–17 21 16.8 19 2 43 17.0 24 2
use the commercially available version without mod- 18–19 23 19.0 21 2 30 18.8 30 0
ifications to establish norms for dexterity in Korean *Right- or left-hand dominance.
elementary school age children.5 However, the age
range in both studies was limited: 5–10 years in the
Smith et al.2 study and 7–12 years in the Yim et al.5
et al.8 was followed in this study. The pegboard was
study. Other studies have shown that dexterity con-
centered in front of the subject with the container
tinues to improve during adolescence,6 and it has
side on the same side as the hand being tested. The
been suggested that adolescents do not have similar
dominant hand was tested first. Subjects completed
dexterity to adults.7 Because the adult norms on the
one practice trial followed by the actual timed test for
9-HPT start at age 20,8 there is no information regard-
each hand. The instructions used were the same as
ing the performance of children between 12 and
those used by Mathiowetz et al. For the nondominant
20 years of age on this test. Therefore, the purpose
hand, the pegboard was turned so that the container
of this study was to develop norms for school-age
was on the same side as the nondominant hand.
children as measured using the 9-HPT across a wider
Interrater reliability was established by having the
age range, and to provide initial normative data for
examiners simultaneously time 20 subjects. Intraclass
4–19-year-old children.
correlations were 0.98 for the dominant hand and
0.96 for the nondominant hand.
MATERIALS AND METHODS
RESULTS
Participants
Four hundred and six children from ages 4–19 The means and standard deviations according to
years were tested (193 males and 213 females) age, gender, and dominance are shown in Table 2.
(Table 1). The majority of the participants (n = 369) Data are presented in two-year age intervals.
were right handed and 47 participants were left A mixed model analysis of variance was calculated
handed, which reflects the 10–15% estimate of left- to examine differences in dexterity times with
handed people in the general population.9 Handed- age and gender for both the dominant and non-
ness was identified by asking the participant or dominant hands. There was a main effect for age
parent which hand was used for writing/drawing. (F7, 372 = 140.95, p , .0001), gender (F1, 372 = 12.50,
Children with a parent-reported or self-reported his- p , .0005), and dominance (F1, 372 = 85.51, p ,
tory of neurologic, orthopaedic disability or with a .0001). Speed of dexterity improved with age. Male
special education classification were excluded from and female scores for each age group were combined
the study. Twenty-one percent of the participants were and post hoc analyses using the least significant
Hispanic, 2% were African American, and 1% were difference test were calculated to determine which
Native American; 86% were from urban areas. The ages differed in dexterity. For the dominant hand
sample represented diverse socioeconomic groups (Table 3), times for children 4–9 years old were
surrounding a large Western metropolitan area. significantly slower than each other and slower
than children older than 10 years of age. Times for
Procedure the 10–11-year-old children were similar to the 12–13
and 14–15 year olds, but were significantly slower
After obtaining informed consent from the partic- than children aged 16 years and older. Times for the
ipant and/or parent, participants were tested indi- 12–13-year-old children were similar to the 14–15
vidually by one member of the research team in a year olds but slower than children older than 16 years
quiet location. Children were tested at a desk and of age. The 14–15-year-old children had times similar
chair of appropriate height with their feet supported to the children older than 16 years of age. For the
on the floor. The procedure described by Mathiowetz nondominant hand (Table 3), times for children

July–September 2005 349

 TABLE 2. Mean Completion Time in Seconds by Age, Sex, and Hand Dominance
Males Females
Age range
(yr) N Dominant SD Nondominant SD N Dominant SD Nondominant SD
4–5 27 29.8 3.8 34.5 5.9 21 30.2 6.3 33.2 6.2
6–7 25 25.5 6.0 28.5 6.6 23 22.5 2.3 25.9 5.2
8–9 23 19.9 3.9 21.7 4.3 26 18.7 1.9 21.2 3.2
10–11 24 18.9 4.1 20.2 3.3 21 16.7 3.4 19.0 3.1
12–13 25 18.0 2.5 18.4 2.6 24 17.1 1.8 18.1 2.2
14–15 25 18.0 2.7 18.6 1.8 25 16.8 2.4 18.1 1.8
16–17 21 16.9 2.0 17.1 2.4 43 15.8 1.9 17.1 1.8
18–19 23 16.1 1.6 16.7 1.2 30 16.1 2.1 17.4 2.0

between the ages of 4–9 years were significantly children older than 12 years. In addition, the scores
different from each other and significantly slower for our 16–19-year-old children were not significantly
than children older than 10 years of age. Times for different from scores for the norms reported by
children between the ages of 10 and 15 years were Mathiowetz et al.8 for adults aged 20–29 years (one-
similar, and times for children older than 12 years of sample t-tests; all p-values were greater than 0.44).
age were similar. Participants performed faster with Thus, the adult norms for the 9-HPT should not be
the dominant hand than with the nondominant used for the dominant hand in children younger than
hand. Females performed faster than the males. The 16 years but could be used for the nondominant hand
only significant interaction was Dominance 3 Age for children older than 12 years of age.
(F7, 372 = 5.53, p , .0001). As children’s ages in- Participants in our study in the 5–10-year-old
creased, the difference in times between the domi- groups were slightly slower overall than the partic-
nant and nondominant hands decreased. ipants in Smith et al.2 study for both the dominant
and nondominant hands. However, our study used
DISCUSSION the commercially available Sammons, Preston and
Rolyan pegboard that was not adapted with the
The data from this study support the conclu- nonskid surface on the bottom or with the shock-
sions from previous studies. That is, females perform absorbent surface, as was the pegboard in Smith
faster in fine motor dexterity tests than males, and et al.’s study. Times for our participants were slightly
dominant hand scores are faster than nondominant faster than those obtained by Yim et al.,7 who did use
scores.2 Older children were faster than younger a commercial-version pegboard.
children, a finding that is different from findings This study was limited by the use of a convenience
in adults that show dexterity on the 9-HPT actually sample, self-reports to determine dominance, and the
decreases with age.8 This was the first study to ex- absence of medical conditions. Future studies might
amine changes in dexterity on the 9-HPT in children want to address a larger number of children and
older than 12 years of age. Several studies suggested represent greater geographic, ethnic, and socioeco-
that dexterity scores as measured by the Purdue nomic backgrounds.
Pegboard change little in children older than 10 years
of age.10,11 Other studies reported that scores on the CONCLUSION
Purdue Pegboard did improve and that adult norms
should not be used for children older than 10 years of The 9-HPT is commercially available, easy and
age.6,7 Our study also found changes in scores for the quick to administer, portable, and requires minimal
dominant hand in the participants older than 10 years space and equipment. The 9-HPT has been shown to
of age in that the 10–13 year olds were significantly be sensitive to change in adults with neuromuscular
slower and showed greater variance than were chil- and musculoskeletal disorders, and correlates with
dren 16–19 years of age. For the nondominant hand, daily tasks requiring dexterity.12–14 However, at
there were no significant differences in scores in the present time, no studies have used the 9-HPT

TABLE 3. Comparison of Combined Mean Male and Female Times for Dominant and Nondominant Hands
Age (yr)
Hand 4–5 6–7 8–9 10–11 12–13 14–15 16–17 18–19
Dominant 30.0a 24.3b 19.4c 18.0d 17.5d,e 17.4d,e,f 16.1f 16.4e,f
Nondominant 34.0a 27.1b 21.5c 19.6d 18.3d,e 18.4d,e 17.5e 17.2e
Note: Means in the same row sharing the same subscript are NOT significantly different at 0.05 in the least significant difference
comparison.

350 JOURNAL OF HAND THERAPY

to document dexterity variability in children with 2. Smith YA, Hong E, Presson C. Normative and validation
pathology, possibly because of the lack of normative studies of the Nine-Hole Peg Test with children. Percept Mot
Skills. 2000;90:823–43.
data. Now that normative data exist from the present 3. Kellor M, Frost J, Silberberg N, Iversen I, Cummings R.
study, the test could be used as a screening tool for Hand strength and dexterity. Am J Occup Ther. 1971;25:77–83.
measuring dexterity in children. Furthermore, the 4. Davis J, Kayser J, Matlin P, Mower S, Tadano P. Nine-Hole
Peg Tests: Are they all the same? Occup Ther Pract. 1999;April:
9-HPT may be particularly useful to measure and
59–61.
monitor hand dexterity in children who have hand 5. Yim SY, Cho JR, Lee IY. Normative data and developmental
injuries, who have undergone surgery, or who have characteristics of hand function for elementary school children
diseases involving the hand (e.g., arthritis). in Suwon area of Korea: grip, pinch and dexterity study.
J Korean Med Sci. 2003;18:552–8.
6. Gardner RA, Broman M. The Purdue Pegboard: normative data
Acknowledgments on 1334 school children. J Clin Child Psychol. 1979;1:157–9.
7. Mathiowetz V, Rogers SL, Dowe-Keval M, Donahoe L,
The authors wish to thank the students for their participa- Rennells C. The Purdue Pegboard: norms for 14-19-year-olds.
tion in this study. They also thank those who helped make Am J Occup Ther. 1986;40:174–9.
this study possible: Lisa Brower, Sheryl Fleck, OTR/L, 8. Mathiowetz V, Weber K, Kashman N, Volland G. Adult norms
Anderson Academy, Albuquerque Academy Day Camp, for the Nine-Hole Peg Test of finger dexterity. Occup Ther J
Belen Public Schools, Butte Public Schools, Campfire Boys Res. 1985;5:24–8.
and Girls, East Mountain Schools, Menaul School, and 9. Porac C, Coren S. Lateral Preferences and Human Behavior.
Serendipity Day School. They thank all the subjects, New York: Springer-Verlag, 1981.
schools, and programs who participated in this study. 10. Costa LD, Scarola M, Rapin I. Purdue Pegboard scores for
This study was completed as partial fulfillment of the normal grammar school children. Percept Mot Skills. 1964;18:
Independent Study Project for the Bachelor of Science 748.
degree for Ms. Kirk and Ms. Markham at the University 11. Siegel M, Hirschhorn B. Adolescent norms for the Purdue
Pegboard test. Pers Guid J. 1958;36:563–5.
of New Mexico Occupational Therapy program.
12. Grant R, Slattery J, Gregor A, Whittle IR. Recording neurolog-
ical impairment in clinical trials of glioma. J Neurooncol. 1994;
19:37–9.
REFERENCES 13. Sterr A, Frievogel S, Voss A. Exploring a repetitive training
regime for upper limb hemiparesis in an in-patient setting: a
1. McHale K, Cermak SK. Fine motor activities in elemen- report on three case studies. Brain Inj. 2002;16:1093–107.
tary school: preliminary findings and provisional implications 14. Backman C, Cork S, Gibson D, Parsons J. Assessment of hand
for children with fine motor problems. Am J Occup Ther. 1992; function: the relationship between pegboard dexterity and
46:898–903. applied dexterity. Can J Occup Ther. 1992;59:208–13.

July–September 2005 351