Maley, Claire (2009) Young children's human figure drawings: an investigation using the Goodenough-Harris Drawing Test and the Rasch Model for measurement. PhD thesis, James Cook University.

Preview

PDF (Thesis) Download (4MB) | Preview

Abstract

This study investigates young children’s human figure drawings using the Goodenough- Harris Drawing Test (GHDT) and the Rasch model for measurement. Applying Constructivist and Latent Trait theoretical perspectives the aim, in particular, was to examine the psychometric properties of the test and the human figure drawings in general, and to investigate whether a more culturally, socially and educationally relevant prototype Human Figure Drawing Continuum (HFDC) could be constructed.

The Goodenough-Harris Drawing Test (GHDT) is a non-verbal assessment of young children’s levels of intellectual maturity which are inferred from the detail and concepts included in human figure drawings. Originally known as the Goodenough Draw-a-Man Test (GDAMT; Goodenough, 1926a), Harris revised and extended the test to include the Draw-a-Woman (DAW), Self-Portrait-Man (SPM; for male children), and Self-Portrait-Woman (SPW; for female children) sub-tests. However, these sub-tests were added to the GDAMT despite there being no empirical evidence indicating that a single drawing of a man was actually insufficient for the task of inferring children’s levels of intellectual development. Similarly, there is currently no empirical evidence that verifies the effectiveness of the tripled data collection load which gathers three human figure drawings from young children (that is, a DAM, DAW and a SPM or SPW drawing). Whilst Goodenough (1926a) and Harris (1963) established superficially the validity and reliability of the GHDT, the test remains unexamined from a modern test theory perspective.

The sample in this study comprised 107 children from years prep to five attending a P-12 school in North Queensland. Data collection consisted of administering the GHDT to the children either individually or in small groups across three phases, each approximately six months apart. The drawings were examined and scored in accordance with the GHDT scoring guides. Results from the Rasch analyses were then used to produce drawing development graphs and common linking plots which displayed visually what the Rasch model detected statistically.

The results indicated that the GHDT was apt for Rasch analysis and the measurement scales produced did not breach the unidimensionality requirement or the other Rasch model expectations. Furthermore, the measurement scales adequately summarised the children’s human figure drawings with very few item and case performances detected as misfitting. The mean person measures indicated that the selfportrait (SPM and SPW) sub-tests were better targeted to the sample of children than the DAM or DAW sub-tests. However, the person mean raw scores indicated that children received less credit for their self-portraits than they did for their drawings of men or women. This suggests that the scoring guides, designed to evaluate drawings of men and women, were significantly less sensitive to drawings of children.

Examination of the development of the children’s DAM and DAW drawings over the approximate twelve month data collection period indicated that the drawings developed in alignment with Piaget’s theory (1956, 1971). However, the Rasch analysis results indicated that drawings of women developed more erratically than those of men.

Interestingly, Rasch output indicated that Harris’s inclusion of the DAW and SPM/SPW components contributed little additional information beyond what can be inferred through the original DAM sub-test. Moreover, the common person linking plots revealed that some boys were disadvantaged by Harris’s inclusion of the DAW sub-test. A common person linking plot including the measures produced by the Rasch analysis of the 50 common items across the DAM and DAW sub-tests suggested that the items exclusive to drawings of men and women were somewhat redundant. Moreover, common person linking plots which compared the researcher’s prototype Human Figure Drawing Continuum (HFDC) and each of the GHDT sub-tests indicated that the 45 item HFDC was just as effective in evaluating the young children’s human figure drawings as the 217 item GHDT (i.e. DAM, DAW and a SPM/SPW sub-test). Furthermore, the HFDC is considered to be more user-friendly, faster to administer and score, and it empowers young children to self-select the type of human figure drawing they would like to complete (i.e. drawings of men, women, boys or girls). Facilitating young children in selecting their own type of human figure drawing aligns more closely to the foundations of early childhood education in Australia which is child-centred, flexible, and play-based.

In addition to the findings described above, a modified drawing booklet and scoring guide were devised for use with the draft 45 item HFDC. Both the drawing booklet and scoring guide are based upon that used for the GHDT, however, modifications were made to bring these in line with current educational, societal and cultural expectations.

Overall, this research confirmed that young children’s human figure drawings can be immensely useful to teachers, psychologists, parents and other interested parties; furthermore, close observation of these drawings can reveal much about the nature of young children’s development.

Item ID:	11382
Item Type:	Thesis (PhD)
Keywords:	children’s development, drawing ability, human figures, psychometric tests of children, intellectual development of children, Goodenough-Harris Drawing Test, Rasch Model for measurement
Copyright Information:	Copyright © 2009 Claire Maley
Date Deposited:	26 May 2010 23:18
FoR Codes:	13 EDUCATION > 1301 Education Systems > 130102 Early Childhood Education (excl Maori) @ 50% 13 EDUCATION > 1301 Education Systems > 130105 Primary Education (excl Maori) @ 50%
SEO Codes:	93 EDUCATION AND TRAINING > 9301 Learner and Learning > 930103 Learner Development @ 100%
Downloads:	Total: 6683 Last 12 Months: 13
	More Statistics