Primary ovarian insufficiency quality of life scale

Assignment 3 Rubric and Answers - Tomita Question Answer Points

1 Is the study population identified and described? Are eligibility criteria specified? Are the sample selection procedures clearly delineated?

Study Population Older patients with heart failure (HF) 5 Eligibility criteria Inclusion criteria: individuals who: (1) were living at home and

60 years of age or older; (2) had at least one emergency room visit or overnight hospital admission due to HF in the past year; and (3) met the New York Heart Association Classification of Heart Failure for Class II or Class III. Exclusion criteria were individuals who: (1) were scheduled for any kind of heart surgery and/or (2) had cognitive impairment (one point for each criterion)

5

Sample selection procedures Patients were selected from three hospitals and two health

insurance companies. The selection procedures were not clearly delineated.

5

2 Do the sample and population specifications support an inference of construct validity with regard to the population construct?

Heart failure-specific symptoms were measured using the Congestive Heart Failure Questionnaire which has good reproducibility and its responsiveness is similar to other measures.

5

3 What type of sampling plan was used? Would an alternative sampling plan have been preferable? Was the sampling plan one that could be expected to yield a representative sample?

Sampling plan Convenience sampling 5 Alternative sampling plan Yes. A probability sampling plan (e.g., random, cluster) would

have yielded a more representative sample. 5

Sampling plan yielding representative sample No. A convenience sample does not usually yield a

representative sample 5

4 If sampling was stratified, was a useful stratification variable selected? If a consecutive sample was used, was the time period long enough to address seasonal or temporal variation?

Stratification Sampling was not stratified. 5 Consecutive sampling No consecutive sampling was used. 5

5 How were people recruited into the sample? Does the method suggest potential biases?

Recruitment Recruitment methods were not clear in the paper. 5

Potential biases Given that this is a convenience sample, the sample may not

be representative. Therefore, we may have systematic error or bias.

5

6 Did some factor other than the sampling plan affect the representativeness of the sample?

Non-response: Figure 2 shows that a large number of the people contacted (129 of 262) did not respond. We don't know if these individuals were different from those who responded.

5

7 Are possible sample biases or weaknesses identified by the researchers themselves?

Yes.

Small sample size 5 No geographic diversity 5 No ethnic diversity 5

8 Are key characteristics of the sample described (e. g., mean age, percent female)?

Yes. Table 1 reports baseline demographic and health characteristics for control and treatment groups. For control group: mean age is 77.5 years; 62.5% female; 87.5% Caucasian; 54.2% 12 years of schooling or less, 37.5% living alone. Treatment group: mean age is 74.2 years; 75% female; 81.3% Caucasian; 56.3% 12 years of schooling or less, 50% living alone.

10

9 Is the sample size sufficiently large to support statistical conclusion validity? Was the sample size justified on the basis of a power analysis or other rationale?

Statistical conclusion validity Statistical conclusion validity refers to the degree to which

inferences about relationships from statistical data analysis are correct. The small sample size did impact the statistical significance of the between-group comparisons (no significant differences were found).

4

Power analysis The authors mention that the sample size was determined by a

power analysis to achieve 80% of power at a significance level of 5%. They also accounted for the high attrition rate expected for the control group by increasing the size of that group.

6

10 Does the sample support inferences about external validity? To whom can the study results reasonably be generalized?

External validity Given the small sample size, the external validity - which refers

to the generalizability of the results - is expected to be low. 5

Generalizability Study results can reasonably be generalized to older patients

who are similar in demographic characteristics to the sample (e.g., primarily Caucasian), who live in the Western New York area, and who are NYHA Class II or III.

5