Article Critique
Jessica Foley (Hingle)
April 2, 2008
Indicate whether you would classify the study as experimental or non-experimental. What is the basis for your decision?
The article entitled "The Effect of Graphic Format on the Interpretation of Quantitative Data" details an experimental study of the effect on recall of presenting numerical data in several different chart formats. This study is experimental because the researcher systematically varies the independent variable, in this case the chart format, in order to determine the effects of the variation.
Indicate whether the review of prior research provided a compelling rationale for the need for the study.
Several studies and recommendations concerning the use of graphics in presentations were cited in the article, however it was pointed out that the studies did not cover the effect on recall of presenting numerical data in several different chart formats. Seeing as a wide range of chart formats is available to the general public, there is good rationale for the need for the study. The findings could be used by a large number of people.
Identify the research question(s), and/or hypotheses addressed.
The article details two main research questions:
1)What is the effect of graphic format (bar, line, table, and line-table graphs) on the interpretation of numerical data?
2)What is the impact of cognitive style (field dependant or field independent) on the ability of the subjects to recall information in any particular chart format (bar, line, table, and line-table graphs)?
There was not a distinctly stated hypothesis for either question, but I would make the assumption that the researcher believed that one of the chart formats would prove superior to the others.
Comment on the adequacy of the sample as being representative of the implied population of interest.
While the sample did indeed include students, many of which are frequently asked to interpret charts, it limited the sample to two very specific types of students: Education and Psychology students currently enrolled in undergraduate studies. This is a convenience sample and not representative of the broad range of individuals who deal with data in the real world. The sample does not address the elementary, secondary, graduate, or vocational levels of individuals. Had the research question stated that the study was to determine the effects of graphic format on the interpretation of numerical data by undergraduate Education and Psychology students, then the sample would be considered more adequate. Additionally, the 96 students were divided into 4 groups, resulting in 24 students per group, which is below the recommended number of 30.
How would you judge the internal validity of the design? What threats to internal validity can you identify?
The following threats to internal validity can be identified from this article:
1)Subject Characteristics: Because this is a convenience sample, the selected group may contain specific characteristics, other than the intended treatment, that could effect the dependent variable.
2)Location: The testing locations were not addressed in the article. If one group were tested in a room that was better lit, would it cause their scores to be higher?
3)Attitude of the Subjects: If the subjects were made aware that they are participating in a study, it is possible that they gave the charts more stringent attention than they would have in a real-world situation. Did that cause less-effective charts to appear to have a higher success rate?
Are the conclusions warranted and do they hold implications for educational practice?
The conclusions are warranted, but simply draw light to the idea that further research should be conducted. The results cannot be directly applied to educational practice at this time. Hopefully, the researcher, or others, can use this study as a basis for further investigation and eventual application to educational practice.