Fix It
Dr. Massey recently saw a headline on CNN.com that read "Kids Who Are Breastfed Are Smarter." She wanted to conduct a study to determine whether this claim is true. To that end, Dr. Massey, who is affiliated with a university in rural North Dakota, contacted her local hospital to obtain a list of all 172 women who gave birth in 2005. She randomly selected 50 women to take part in the study and 18 agreed to participate. Dr. Massey gave all 18 participants the same survey. There were many questions on the survey, but one was critical to testing her hypothesis. It was phrased to determine whether the mothers fed their babies only breast milk for the first year or fed their babies only formula for the first year. This question was then used to divide her participants into two treatment conditions: Breastfed or Formula-fed. Dr. Massey was very pleased to find that exactly half of the women (n = 9) breastfed their babies for one year, and exactly half (n = 9) formula-fed their babies for one year.
To assess intelligence in the children (now 10 years old), Dr. Massey administered an IQ test. Her university does not have access to the most commonly used IQ test so she developed one on her own that she believes to be a comparable assessment of intelligence. Dr. Massey finds that the mean IQ of the breastfed children was 108.5, and the mean IQ of the formula-fed children was 101.0. Based on this information she concludes that breastfeeding does, in fact, causes an increase in intelligence.
Dr. Massey developed her research project based on a CNN headline. What do you think of that? What concerns and limitations do you see with Dr. Massey's study? How might you fix them?

Question

Quizplus · Accepted Answer

The Answer is Research questions may be based on a variety of sources (Ref 2-1, p. 24), including newspaper articles. Dr. Massey's research question is a good one because it is empirical; the question can be subjected to scientific examination (Ref 2-1, p. 25). By obtaining participants from a hospital close to her, Dr. Massey is obtaining a convenience sample. Convenience samples are easily accessible to researchers and most commonly used in research, but are usually not good representations of the population of interest (Ref 4-1, p. 104). Dr. Massey may consider recruiting participants from hospitals across the country to get a more representative sample of mothers. Another issue is nonresponse bias. The fact that less than half of the randomly selected participants agreed to complete the survey may indicate a systematic difference between the children whose mothers participated and the children whose mothers did not (Ref 4-1, p. 106). Dr. Massey measured intelligence using an instrument she developed. Although she believes it is a good assessment of IQ she cannot conclude such until she provides evidence of its reliability and validity (Ref 4-1, p. 98). Last, Dr. Massey concludes breastfeeding causes intelligence but is basing this on mean IQ values only. This is problematic for two reasons. First, she must conduct an inferential statistical test to determine whether, in fact, the mean scores differ due to whether the children were breastfed, or to chance alone (Appendix A, p. 485), and second, her research design does not satisfy the requirements for causation. Specifically, to establish causal inference one must establish covariation, temporal precedence, and rule out all other possible explanations. The independent variable (food type) is not controlled by the researcher. It is not known for how long kids were fed breast milk or formula or what other factors may be present in their environment that might affect intelligence. By design, Dr. Massey cannot make a statement of causation. Although it is a two-group design, because the independent variable lacks proper control it is not a True experiment (Ref 8-1, p. 241).

Fix It Dr. Massey Recently Saw a Headline on CNN.com That Read