Sample-Size Calculation and Sampling Technique

Calculation of magnitude of the outcome variable (functional disability), a continuous variable and scored out of 100, used SD (variance) from a similar previous study. To calculate sample size, a single population–proportion formula of a continuous variable (n = Z [α/2)² σ²/d²]), and the SD from the previous study (σ²=0.18) was used,23 where n =is the required sample size — n = Z (α/2)² σ²/d², σ² = 0.18:

(1.96) (1.96) (0.18) (0.82)/(0.05) (0.05) = 226

and Z the reliability coefficient at 95% confidence (1.96)

W (margin of error) = 0.05

N (nonresponse rate 10%) = 26

As such the sample size was 226+26 =252

Since the population was <10,000, a correction formula was used: nf = n/1 + n/N = 252/1+252/3,500=235

We used consecutive sampling. All respondents who came for psychiatry-outpatient visits during the data-collection period and fulfilled the inclusion criteria were selected sequentially, based on their order of arrival at the hospital.