The Efficacy of Early Intervention Programs: A Meta-Analysis

NICHCY’s Structured Abstract 52 describes the following:

Title| The Efficacy of Early Intervention Programs: A Meta-Analysis

Author | Casto, G., & Mastropieri, M.A.

Source Exceptional Children, 52(5), 417-424.

Year Published | 1986

Meta-analysis techniques were used to statistically combine findings from 74 primary research studies investigating the effectiveness of early intervention with preschoolers with disabilities. Findings indicate that (a) early intervention with children who have disabilities produces a positive sizeable effect size; (b) longer, more intense programs are associated with effectiveness; and (c) there is little support for commonly held beliefs about early intervention, such as the importance of age at start and the degree of parental involvement.

Previous research has shown that early intervention programs are, in general, effective. Earlier reviewers agree that 4 variables are associated with effective interventions: parental involvement, age of the child when he or she begins early intervention (referred to as “age at start”), degree of structure, and the duration and intensity of the services. The general belief is that more parental involvement is beneficial, that it is better to start interventions earlier; and that highly structured, longer, and more intense programs are better. However, these conclusions have been drawn from studies that involved only at-risk or disadvantaged children and have been based on a limited number of studies. The intent of this meta-analysis was to achieve a more comprehensive review of the early intervention research literature that focused only on children with disabilities and to include a larger sample of primary research studies.

Research Questions
Is early intervention effective?

Research Design

  • Number of Studies Included | 74
  • Number of Subjects | Not reported
  • Years Spanned | 1937-1984

Research Subjects
The subjects in this meta-analysis were categorized according to the following disabilities:

  • Mental retardation, 44%
  • Combination of disabilities

Age/Grade of Subjects
Birth to 5 years old

Specified Disability
The majority of the children (44%) were identified as having mental retardation.

Early intervention

Duration of Intervention
Not reported

In general, early intervention programs were found to have immediate benefits for populations with disabilities. These results were evident over a variety of outcome variables, including IQ and motor, language, and academic achievement.

Data on the 4 most-often cited variables associated with intervention effectiveness were also analyzed: parental involvement, age at start, degree of structure, and duration/ intensity.  The following findings emerged:

  1. Parent involvement: The analysis showed that, even though parents can be effective interveners, they are probably not essential to intervention success, and those intervention programs that utilize parents are not more effective that those that do not.
  2. Age at start: There was minimal data to support the belief that “earlier is better” in starting intervention programs.
  3. Degree of structure: Little difference in effectiveness appeared in programs with various degrees of structure. There was a tendency that favored the more structured programs, but the data were inconclusive.
  4. Intensity /duration: This variable proved an important one for children with disabilities. Longer, more intense programs were associated with intervention effectiveness for these children.
  5. Insufficient data existed to draw conclusions about the maintenance of benefits for the populations studied. For disadvantaged populations generally, the immediate benefits of early intervention declined rapidly and largely disappeared after 60 months.

Combined Effects Size
The average effect size* of overall findings from all studies = 0.68.

The effect size, when only studies with good quality internal validity indices were considered, was 0.40. When only studies that met both the criteria of having good quality internal validity* indices and providing immediate posttests were analyzed the effect size was 0.43.

The adjusted average effect sizes for parent participation that was characterized as “minor or not at all” = 0.72. For parent participation characterized as “major or sole,” the adjusted average effect sizes = 0.59.

The adjusted average effect sizes for different ages at which intervention began were:

  • 0.59 (0-6 mos.)
  • 0.47 (6-18 mos.)
  • 0.54 (18-36 mos.)
  • 1.06 (36-48 mos.)
  • 0.70 (48-66 mos.)

The adjusted average effect sizes for different levels of structure in the intervention curriculum were: 0.88 for very structured, 0.79 for somewhat structured and 0.76 for little or no structured.

Finally, the average of adjusted effect sizes for interventions of different intensities were: 0.45 for 500 total hours.

The data support the idea that early intervention programs provide immediate benefits. However, researchers should be careful when assuring that early intervention programs should start as early as possible and that they should involve parents as much as possible, because there is little evidence to support these claims. It seems true that highly structured programs are better than less structured programs, but only for disadvantaged populations. Future research studies in efficacy areas should have more rigorous design requirements and should utilize broader-based outcome measures.


* Meta-Analysis | A widely-used research method in which (1) a systematic and reproducible search strategy is used to find as many studies as possible that address a given topic; (2) clear criterion are presented for inclusion/exclusion of individual studies into a larger analysis; and (3) results of included studies are statistically combined to determine an overall effect (effect size) of one variable on another.

Back to top