The article, “Effects of WIC and Food Stamp Program Participation on Child Outcomes,” written by researchers Bong Joo Lee and Lucy Mackey-Bilaver, provides the results of a six-year causal-comparative research study. The study’s purpose is to determine the impact of WIC and Food Stamp Program (FSP) participation on child welfare (outcomes). The article’s title identifies child outcomes as the dependent variable and the independent variables as WIC and FSP. Regression analysis or Ordinary Least Squares, mentioned in the article’s Abstract and the Introduction and Data and Variables’ sections, is the statistical tool utilized to test the data.

Variable AnalysisThe dependent variable, child outcomes, falls into two categories: 1.) the incidence or occurrence of children’s health problems and 2.) the incidence or occurrence of child maltreatment (reports of child abuse and/or neglect). The dependent variable is a discrete variable, meaning the researcher received a count of child outcome occurrences, and its measurement level is ratio.The independent variables are WIC and FSP participation. These are discrete variables, meaning the researcher received a count of someone’s participation in either program, and their measurement level is ratio.

HypothesesThere are a series of clearly stated hypotheses analyzed in this research study. They are as follows:Null: WIC and FSP participation have no affect on the incidence of children’s health problems.0H : µ1 = µ2Alternative: WIC and FSP participation will reduce the incidence of young children’s health problems (directional).1 :H : µ2 < µ1Alternative: WIC and FSP participation will affect the incidence of young children’s health problems (non-directional).1 :H : µ2 ? µ1Null: WIC and FSP participation have no affect on the incidence of child abuse/maltreatment.0H : µ1 = µ2Alternative: WIC and FSP participation will reduce the incidence of child abuse/maltreatment (directional). 1 :H : µ2 < µ1Alternative: WIC and FSP participation will affect the incidence of child abuse/maltreatment (non-directional).

1 :H : µ2 ? µ1Sample AnalysisThe sample size gathered is large enough at 252,246 because it is greater than thirty. This follows the Central Limit Theorem, which states that the sampling distribution of the mean can be approximated closely with a normal distribution. I, however, could not calculate standard error of the mean because the authors did not provide the overall standard deviation for the outcomes or incidence of health problems/child abuse.

The authors provided the statistical significance for the independent variables. The author did, however, provide the percentage of the standard deviation, sorted by demographic characteristic (i.e., race, poverty level). If we were calculating the Z statistic, the rejection region would be less than -1.645 at the 5% significance level for a one-sided (directional) hypothesis; ± 1.

96 at the 5% significance level for a two-sided (non-directional) hypothesis.Results and ConclusionsAppropriate statistical techniques are used in the research reported because regression analysis (OLS) is the tool used to test the effect of an independent variable on a dependent variable. Since the level of measurement for both the independent variables and the dependent variables is ratio (meaning the absence of incidence occurs at the zero point), OLS is an appropriate technique. Charts were effectively used to illustrate the percentage of occurrence by demographic group. A line graph with scatter points would have been helpful to show the spread of the data over time, since this is a longitudinal study that occurred over a five-year period. Bar graphs would have provided additional clarification of the occurrence of the dependent variables as a result of the independent variables.

Review of the study’s conclusions suggests that the researchers clearly related the results to the hypotheses. The researchers provided numerous examples of the level of statistical significance (? = .05), which suggest that the independent variables had a statistically significant impact on the dependent variables. In each instance, the significance level was < .05, indicating a rejection of the null hypotheses that the independent variables would have no impact on the dependent variables. Participation in WIC and FSP (both jointly, multiple regression, or separately, single regression) had a significant impact on the occurrence of child outcomes (health problems and abuse). Thus, the findings suggest that WIC and FSP reduce the level of occurrence of child outcomes (health problems and abuse).

It, however, would have been interesting and valuable if the researchers would have provided statistics indicating how much the independent variables caused changes (variation) in the dependent variables and what other variables (not included in the study) may have affected the outcomes.