Study presents unadjusted program and comparison group means (columns 5 and 6). Tests of statistical significance apply to adjusted results in column 7 (means difference or odds ratio) (Minkovitz, et al., 2001).
The standard errors in Minkovitz et al. (2001) are adjusted for clustering by site (Minkovitz, et al., 2001).
NCAST total score
HOME total score
HOME positive involvement
HOME lack of negativity
HOME cognitive stimulation
NCAST Total Score Mean = 47.61 (SE = 0.42)
P/CIS general mean = 3.71 (SE = 0.05)
P/CIS appropriateness mean = 3.47 (SE = 0.03)
HOME total score mean = 39.07 (SE = 0.26)
HOME positive involvement mean = 5.69 (SE = 0.05)
HOME lack of negativity mean = 4.88 (SE = 0.12)
HOME cognitive stimulation mean = 6.78
NCAST Total Score mean = 48.58 (SE = 0.47)
P/CIS general mean = 3.85 (SE = 0.05)
P/CIS appropriateness mean = 3.59 (SE = 0.03)
HOME total score mean = 39.65 (SE = 0.29)
HOME positive involvement mean = 5.70 (SE = 0.05)
HOME lack of negativity mean = 5.03 (SE = 0.13)
HOME cognitive stimulation mean = 6.72 (SE = 0.10)
NCAST Total Score MD = -0.97
P/CIS general MD = -0.14
P/CIS appropriateness MD = -0.12
HOME total score MD = -0.58
HOME positive involvement MD = -0.01
HOME lack of negativity MD = -0.15
HOME cognitive stimulation MD = 0.06
The standard errors in Caughy et al. (2004) are adjusted for making multiple comparisons (Caughy, et al., 2004).
Study presents unadjusted program and comparison group means (columns 5 and 6). Tests of statistical significance apply to adjusted results in column 7 (means difference or odds ratio) (Minkovitz, et al., 2007).
This table shows a study’s results by outcome measure. Rows for primary outcomes
are shaded. Favorable outcomes are listed in green and unfavorable or ambiguous
outcomes are listed in red.
This table describes in more detail the outcome measures used by a study. Rows for
primary outcomes are shaded.
Primary outcome measure: For the HomVEE review, an outcome measured
through direct observation, direct assessment, or administrative data; or self-reported
data collected using a standardized (normed) instrument.
Secondary outcome measure: For the HomVEE review, most self-reported
data, excluding self-reports based on a standardized (normed) instrument.
For a uniform approach, HomVEE calculates mean differences as the
program group mean minus the comparison group mean. In some studies, authors did
this calculation in the opposite direction.
The odds ratio is the odds of an event occurring in one group divided
by the odds of an event occurring in another group. If the odds of an event occurring
in group A equals 3 and the odds of an event occurring in group B equals 2, the
odds ratio equals 1.5. In other words, there is a 1.5 times greater likelihood the
event will occur in group A than in group B.
HomVEE reports author-calculated effect sizes when available, as labeled in the
table. If the author did not report effect sizes, but sufficient information was
available, HomVEE calculated effect sizes.
The HomVEE review calculates effect sizes using the same computations as the
What Works Clearinghouse (WWC). The WWC, established by the Institute for
Education Sciences in the U.S. Department of Education, reviews education research.
For more information, please read
WWC effect size calculations.
Statistical significance is an indication of the probability that the observed finding
was obtained by chance (when there is not a real relationship in the population).
If the p-value is equal to or less than a predetermined cutoff (in the
HomVEE review, 0.05), then the finding is considered statistically significant or
probably not obtained by chance.
Home | Research | Outcomes | What's New | Publications | Review Process | About Us | Help