Skip to main content
. Author manuscript; available in PMC: 2011 Jan 1.
Published in final edited form as: J Appl Dev Psychol. 2010 Sep;31(5):351–361. doi: 10.1016/j.appdev.2010.06.004

Table 1.

Means, standard deviations, ranges, and skewness for predictor variables before combination.

Mean SD Maximum Maximum Skewness*
Neighborhood measures:
  Rater report of risk 6.97 1.55 4 13 .81
  Mother report of problems 9.83 2.74 6 24 1.27
  Mother report of social org. 17.06 4.46 7 28 −.03
Temperament measures:
  Child report
    Fearfulness 19.87 4.83 6 30 −.23
    Irritability 26.14 5.62 8 40 −.12
    Impulsivity 20.11 4.80 7 33 .09
  Mother report
    Fearfulness 17.59 3.94 8 29 −.09
    Irritability 26.81 5.43 11 40 −.05
    Impulsivity 22.35 4.33 11 33 −.02
Adjustment measures:
  Child report
    Depression/anxiety 31.88 12.51 7 58 −.22
    Conduct problems 4.41 4.05 0 31 1.90
    Social competence 61.96 9.16 20 80 −.56
  Mother report
    Internalizing problems 9.31 6.34 0 33 .93
    Externalizing problems 10.71 7.01 0 33 .86
    Social competence 42.22 7.04 18 56 .14
*

All skew errors ranged from .136 to .138 (all rounded to .14). Several of the measures' skewness statistics were significant (>1.96). However, large samples give rise to small standard errors, and significant values arise from even small deviations from normality, thus, this criterion can be misleading. Skewness statistics were markedly reduced in the combined measures. Further, visual inspection of the histograms of the combined variables used in analyses suggested that all variables approximated normal distributions.