A study recently published in the San Jose Mercury News suggests the parents of obese children do not perceive their children as obese. An analysis of this study, its methods, and its findings is an important exercise in understanding the meaning and relevance of all research. The ability to read research critically and understand how it was generated allows us to identify possible design flaws or to realize the validity of its conclusions and make appropriate use of the data.
The study was conducted through an internet research firm for the University of Michigan. Its goal was to determine the percentage of parents who realized their children are obese and to compare it to the percentage of parents that do not realize their children are obese. The participants were selected by simple random sampling and were considered a representative sample of American parents. The sample included 2060 respondents (Runk, 2007).
The data appeared to be collected through internet surveys. The study was observational and measured continuous data sets including the height, weight, age, and gender of the participant’s children (Bennett, Briggs, & Triola, 2003). A body mass index greater than or equal to the 95th percentile in comparison to children the same age and gender was the criteria used for obesity in this study. The data from these children was then compared the national percentage of children considered obese by the same standards. Qualitative data pertaining to whether or not the parents thought their children were “slightly overweight, very overweight, or about right” was also collected (Runk, 2007).
Although it was not specifically stated, I hypothesize the children studied were broken down into bins by age and gender. The first bin comprised of obese girls six to 11 years old and the second bin obese boys six to11 years old. For the 12 to 17 year old group the first bin included obese girls 12 to 17 years old and the second bin included obese boys 12 to 17 years old. I further hypothesize the parents of children in each group were binned according to their response to the qualitative assessment of their child’s weight. For each age group the first bin included parents who answered ” very overweight,” the second bin included parents who answered “slightly overweight,” and the third bin included parents who answered “about normal” (Bennett, Briggs, & Triola, 2003).
The findings in the six to 11 year old group revealed 15% of the children in this age group met the criteria for obesity. This finding was not significantly different from the national figure which reports 17% of all children in the United States are obese by the standard of this study. Thirteen percent of the parents of obese children in this age group categorized their children as “very overweight,” 37% classified their children as “slightly overweight,” and 43% reported their children were about right” (Runk, 2007).
The findings in the 12 to 17 year old group revealed 10% of children in this age group met the criteria for obesity. This finding is significantly lower than the national figure for obese children. Thirty one percent of the parents of obese children in this age group reported their children as being “very overweight,” 56% reported their children as “slightly overweight,” and 11% reported their children were “about right” (Runk, 2007).
Researchers concluded both age groups under reported the incidence and severity of obesity when compared to the national statistic stating 17% of all children meet the criteria for obesity outlined in this study (Runk, 2007). I agree the severity of obesity was greatly under reported in both groups. However, I disagree with the assertion the incidence of obesity was under reported in the six to 11 year old group. I believe the difference between 15% and 17% could easily be a coincidence. It may also have resulted because the internet was used to collect data and poor children are more likely to be obese and less likely to have internet access (Vieweg, Johnston, Fernandez & Pandurangi, 2007).
I do agree that obesity seemed to be considerably under reported in the 12 to 17 year old group. A statistically significant difference (about 7%) occurred between that age groups 10% incidence and the 17% national incidence of childhood obesity (Runk, 2007). Such a large difference is unlikely to be a coincidence and supports the theory that obesity was under reported in this age group or confounding was present (Bennett, Briggs, & Triola, 2003).
Critical analysis of this data reveals many strengths and a few significant weaknesses in the design and implementation of this study. The goal is clearly stated, to determine the percentage of parents who realized their children are obese and to compare it to the percentage of parents that do not realize their children are obese. This goal was clearly accomplished for all the study participants. The source of the study is the University of Michigan which can be considered a reliable, neutral source. The sampling is sufficiently large, but whether or not it is representative of childhood obesity in this country is questionable. A serious problem with the sample exists as a result of using the internet as the setting. High proportions of obese children are socioeconomically disadvantaged and may not have internet access.
The internet setting is likely to account for the apparent under reporting of obese children noted in this study and as such is a probable source of confounding. The criterion for obesity is well defined and could be easily measured in all the subjects, but I remain concerned a significant portion of obese children may have been inadvertently omitted from consideration. In the end, however, I find there is a strong practical use for this data. Healthcare providers are made conscious of the fact that the parents are, more often than not, genuinely unaware their child is obese. This data supports the decision to open a dialogue with parents and offer teaching about the dangers and prevention of childhood obesity the clear course of action.
Bennett, J., Briggs, W., & Triola, M. (2003). Statistical Reasoning for EverydayLife, Second Edition. Upper Saddle River, NJ: Addison Wesley. RetrievedDecember 5, 2007, from University of Phoenix rEsource HCS 438.
Runk, D. (2007, December 24). Parents don’t realize their kids are fat. MercuryNews (San Jose). Retrieved December 28, 2007, fromhttp://www.mercurynews.com/healthandscience/ci_7799918?nclick_check=1.
Vieweg, V., Johnston, Fernandez, A., & Pandurangi. A. (2007). Correlation
between high risk obesity groups and low socioeconomic status in school children. Southern Medical Association. Retrieved January 12, 2008, from University of Phoenix library [EBSCOhost].