Mona Maarse Delahooke (1986) attempted to determine how home schooled and one group (private) of conventionally schooled children compared in terms of social/emotional adjustment and academic achievement. This was the subject of her Ph.D. dissertation research at California School of Professional
Psychology, Los Angeles. She reported that the purpose of her study “was to examine variables cited in the literature as rationales for engaging in home education: the effects of increased parental contact on academic achievement, and on social and psychological functioning” (p. 12).
The researcher developed a theoretical argument based upon “dual process” theories. That is, both familial and peer systems have significant roles in the development of a child’s social competency. Dual process theories, as described by Maarse Delahooke, emphasize “the importance of a secure familial base
from which to explore, as well as the presence of peer interaction, in order to develop social competency as adults” (p. 22). She went on to review literature related to the results of child interactions with peers and adults. Based on the literature and the perceived nature of home schooling, “It was thus expected that home educated children’s decreased school peer experiences and increased adult contact would result in the expression of fewer aggressive social behaviors, and a greater amount of perceived reliance on and support from adults than more traditionally educated children” (p. 39).
Maarse Delahooke also made a theoretical case, based upon the literature, for the idea that increased parental involvement in education leads to higher academic achievement. Thus, she hypothesized that home educated children would score higher academically than the “traditionally” private schooled children.
The researcher took a logical approach by attempting to obtain two comparison groups which were similar except for one variable, the mode of instruction. To the aforementioned extent, it was a causal comparative research design. All families resided in California, and the two groups (28 home educated and 32 private schooled) were very similar in terms of intelligence, age (about 9.1 years), sex ratios, and socioeconomic status. She was careful to explain that it was an exploratory study, and that it did not involve a causal or true experimental design.
The Wide Range Achievement Test-Revised (WRAT-R) was used to assess basic achievement in reading and arithmetic. She cited a report in which the test-retest reliability coefficients for the WRAT-R were .96 for reading and .94 for arithmetic. The test is reportedly “sensitive to developmental changes, as raw score means were noted to increase with age. Further data on the validity of the test, as reported in the WRAT-R manual [sic] include concurrent validity scores … ” (p. 46) ranging from .66 to .87. A short form version of the WISC-R (Weschler) was used to measure and to account for the effects of intelligence. The three I.Q. scales of the WISC-R have reliability coefficients of at least .89, reported the researcher. Concurrent validity coefficients for the WISC-R range from .39 to .82. The Roberts Apperception Test for Children (RATC) was the sociometric instrument used to measure “`…children’s perceptions of common interpersonal situations’ … ” (p. 48). Referring to the RATC manual, Maarse Delahooke reported that with respect to the profile scales (e.g., reliance on others, aggression, and limit setting), “Interrater agreement was reported to range from 85.8% to 92.8% when scored by doctoral-level raters. Additionally, the split-half reliability of the test was reported to range from .44 to .85 for each of the profile scales. The validity properties of the RATC were examined through several different methods, … ” (p. 51). Apparently, the RATC possesses construct validity in that a reported regression analysis significantly discriminated between the scores of two groups of subjects (200 “clinic” children and 200 “well-adjusted” children”). “Thus, the profile scales of the test appeared to be useful in differentiating dimensions of personality, as well as providing indications of adaptive and maladaptive functioning” (p. 52).
All subjects were tested individually in their homes between August 14 and December 3 of 1985. This writer suggests that this was too long a period for optimum control in Maarse Delahooke’s study, especially if there was a bias in terms of which groups were tested later (since potentially more learning could have transpired) or earlier. Ideally, all children should have been tested within a much shorter period.
One-way analysis of covariance was used “to determine if the achievement test scores of home educated children differed significantly (p=.05) from traditionally educated children, while controlling for the effects of intelligence” (p. 52). “T tests were conducted to determine if the means [of the RATC scores] differed significantly between the two groups. Due to the large number of t-tests performed on the RATC scales, a statistical correction was conducted in order to control for an inflated error rate” (p. 53).
The statistical analyses revealed no significant differences between the groups on the Aggression, the Reliance on Others, Support-Other (from whom they perceive support), Limit Setting (who sets limits on them), the family arena of the Interpersonal Matrix, or the Support-Child (self-reliant and assertive behaviors) scales on the RATC. Likewise, there was no significant difference in academic achievement based on type of schooling.
On the other hand, significant differences were found in the “non-family” categories of the Interpersonal Matrix. “This finding suggests that children in the private school exhibited a greater focus on peer interaction, while the home educated children’s primary focus was in the family arena” (p. 83). “…private school subjects appeared to be more influenced by or concerned with peers than the home educated group” (p. 85). The researcher also said that the data may suggest that private school children were more adept at resolving interpersonal problems than those educated at home. A close examination of the data also revealed a trend toward home educated children perceiving a greater amount of supportive behaviors from family as opposed to non-family members. Maarse Delahooke reported that it appears home educated children perceived their parents as primary authority figures more often than did the private school children.
In conclusion, Maarse Delahooke found: 1. Both private and home school groups scored in the “High Average (Bright)” category with respect to intelligence. 2. The two groups did not differ on academic achievement. 3. Neither group performed as high academically as their I.Q. levels would have predicted. 4. Both groups scored in the “well adjusted” range of social/emotional functioning. 5. The two groups ” … did display differences in their perceived social sphere of influence, with the private school group appearing to be more influenced by peers than the home school group” (p. 87). The researcher also suggested that since both groups were involved in peer socialization “(albeit from different sources)” (p. 86), the dual process theories that she had discussed were neither proved nor disproved by her investigation. The researcher pointed out that the study was exploratory in nature, “and attempted to assess general trends in private versus home education” (p. 44).
This study is quite significant in that it carefully controlled a number of variables while allowing the basic mode of instruction (home versus conventional school) to differ. Although a true experimental design was not involved, the causal comparative type of design does lend itself to some inferences. If all things with respect the students were equal except for the mode of instruction they experienced, then one can suggest that home schooling may result in children being less influenced by their peers than is the case for those children in conventional schools. Of course, investigations of this nature are difficult to control perfectly. However, Maarse Delahooke’s study is significant in that she actually attempted to ascertain whether one of the stated objectives of home schooling is being achieved. She has provided one piece of evidence that perhaps one of the objectives (children being less influenced by their peers) of home education is being met.