PUBH 4100 Week 10 Attribution Argument Example

How Much of the Change Belongs to the Program? An Attribution Argument for the Swim Lesson Voucher Evaluation, Weighing Four Alternatives and Setting a Ceiling on Its Own Conclusion

Student Name

Walden University

PUBH 4100: Evaluating Public Health Programs

Instructor Name

Month Day, Year

How Much of the Change Belongs to the Program? An Attribution Argument for the Swim Lesson Voucher Evaluation, Weighing Four Alternatives and Setting a Ceiling on Its Own Conclusion

The Change to Be Explained

Among children who attended six or more lessons, the share meeting the five-element water competence checklist rose from 18.5% at the first lesson to 72.0% at the final lesson, about four weeks later. The question here is how much of that 53.5-point rise can be credited to the lessons. The design has no comparison group, so the answer must come from weighing alternatives against the evidence the evaluation did collect.

Alternative 1: Children Would Have Improved Anyway

The Argument

Children grow and gain coordination over a summer.

The Evidence

The interval between tests was about four weeks. Treading water for a minute and moving 25 yards are learned skills; it is implausible that most children who could not do them would acquire them in four weeks through growth alone. Pre-post designs are weakest when long intervals allow natural change (Harris et al., 2006), and this interval was short.

Verdict

Maturation explains little of the rise.

Alternative 2: Practice Outside the Program

The Argument

Some children may have gained skill from weekend trips to a family pool or a lake beach between sessions.

The Evidence

At the final lesson, parents reported whether their child had swum elsewhere during the session. In the composite data, 31% said yes. Pass rates were 79% among children who had swum elsewhere and 69% among those who had not. Outside swimming is associated with higher pass rates, but most children who passed had not swum elsewhere.

Verdict

Outside practice contributes to some passes but cannot explain most of the change.

Alternative 3: Lenient Scoring

The Argument

Instructors scoring their own students may pass borderline children.

The Evidence

A second instructor from another pool scored 82 randomly selected tests and agreed with the first on 91.5% of results, with a Cohen's kappa of 0.79, commonly read as substantial agreement (McHugh, 2012). Of the seven disagreements, four were cases where the second scorer failed a child the first had passed.

Verdict

Some leniency is possible, perhaps a few percentage points, but scoring does not explain the rise.

Alternative 4: Selection Into Completion

The Argument

The pass rate describes children who attended six or more lessons. Children who dropped out may have been those struggling, leaving stronger swimmers.

The Evidence

First-lesson pass rates were similar among children who later completed six lessons and those who did not, 18.5% and 17%, so completers did not start out stronger. Selection affects whom the finding describes more than whether lessons produced the change among completers.

Verdict

Selection limits generalization to all voucher users but does not explain the rise among completers.

The Language Settled On

Weighing the alternatives, the evaluation concludes: "Most of the increase in water competence among children who completed the program is likely attributable to the lessons, with a smaller contribution from swimming outside the program and possibly from lenient scoring." The word "likely" signals judgment, not proof. The phrase "among children who completed the program" keeps the claim to the group measured.

The Ceiling on the Conclusion

The conclusion cannot rise above this. It cannot say by how many points the program raised competence, because without a comparison group the counterfactual is unknown. It cannot say the program reduced drowning, although other researchers have connected early formal instruction with reduced drowning odds (Brenner et al., 2009). And it cannot extend to the 87% of eligible children who never used a voucher. A waitlist comparison next year would allow a stronger statement.

References

Brenner, R. A., Taneja, G. S., Haynie, D. L., Trumble, A. C., Qian, C., Klinger, R. M., & Klebanoff, M. A. (2009). Association between swimming lessons and drowning in childhood: A case-control study. Archives of Pediatrics & Adolescent Medicine, 163(3), 203-210. https://doi.org/10.1001/archpediatrics.2008.563

Harris, A. D., McGregor, J. C., Perencevich, E. N., Furuno, J. P., Zhu, J., Peterson, D. E., & Finkelstein, J. (2006). The use and interpretation of quasi-experimental studies in medical informatics. Journal of the American Medical Informatics Association, 13(1), 16-23. https://doi.org/10.1197/jamia.M1749

McHugh, M. L. (2012). Interrater reliability: The kappa statistic. Biochemia Medica, 22(3), 276-282. https://doi.org/10.11613/BM.2012.031