Data Selection and Retention - Case A2
As Alan examined his accumulated data, he noted that a number of cell culture plates failed to respond to the hormonal stimulus and that there was considerable variability in the dose- response relationship. His data are represented in Figure 1.
Furthermore, on reexamination, he noted that a number of his gels were not very aesthetic in appearance, yet he was sure that they demonstrated the molecular weight, agonist binding, and subunit characteristics of the receptor .
Alan mentioned his distress to Pam Alden, a fifth-year graduate student, who said, "Why don't you clean up your data? You'll never get the paper published unless you do. We always clean up the data around here. " She then suggested that the four culture points failing to show a response (along the X-axis at the O nanomolar concentration) be dropped because the cells were probably dead. She also pointed out that he might eliminate the top data point at the 45 minute interval as an outlier. She examined the gels and suggested retouching the negatives from which the prints were to be made, including the duplication of one of the nicer gel lanes to replace another that turned out poorly, but showed essentially the same result. "That will greatly improve your chances of publication, " she said. Alan replied, "Maybe I should repeat a few of the experiments or try to improve the culture conditions?" "No," said Pam, "If you're convinced of your results, why go through the time, expense, and uncertainty of more repetitions? You 'II never complete an experiment in time for the abstract, anyhow. " Somewhat dismayed, Alan thanked her and turned back to his work.
- What do you think about Pam's comments on publication practices and her suggestions for "cleaning up" the data?
- How should Alan go about determining which points to include and which to exclude?
- What other course(s) of action would you recommend to Alan?
- Pam's perception about improving the chances of publication by "cleaning up" the data is not uncommon. How might journal editors and reviewers work toward correcting this perception?