• I forgot to include that info, thanks. The significant change in return rate would be 8% in this case (i.e. increasing from 57% in the baseline to 65% post-intervention). Given this, what approach could be used to calculate minimum sample size for the evaluation for alpha=.05 and 80% power?

  • You have a baseline return rate of 57%. In order to determine your sample size for any kind of power and any kind of alpha you will need to decide what constitutes a significant CLINICAL change in return rate. Without this estimate for a change in return rate there is no answer to your question.

  • Andy Gonyea became a registered member 7 hours ago

  • Ina Dickerman became a registered member 9 hours, 9 minutes ago

  • Ursell became a registered member 11 hours, 26 minutes ago

  • Benjamin Snipes became a registered member 11 hours, 32 minutes ago

  • Shauna Rickett became a registered member 12 hours ago

  • Jere Woodriff became a registered member 14 hours, 52 minutes ago

  • Hi, I am planning to do a post-intervention evaluation study and struggling with how to determine the minimum sample size needed. The study is about retention of HIV+ pediatric patients in chronic care after they are diagnosed. The most important variable is percentage of HIV+ children patients who come back to the clinic for treatment one month…[Read more]

  • Vijay Narayan became a registered member 16 hours, 43 minutes ago

  • Juliane Pitts became a registered member 18 hours, 14 minutes ago

  • Ronda Kirkcaldie became a registered member 18 hours, 28 minutes ago

  • Scott Elliott became a registered member 18 hours, 57 minutes ago

  • Akash Swami became a registered member 1 day, 1 hour ago

  • Anthony Tarantino became a registered member 1 day, 18 hours ago

  • Steve Lewin became a registered member 1 day, 21 hours ago

  • @straydog is spot on….If you are using this to justify people, salary, org structures, etc. you need more data or your leaders will not believe your figures.

    I go back to my earlier statement…What is your end goal?

    if it is to show cost per incident then you have to throw out the 264 hours to close number because of all the non value work…[Read more]

  • Dave Perry became a registered member 2 days, 3 hours ago

  • You’re making a false assumption that time to close a task = hours worked when in fact most of this time is non-value-added wait time when nothing is happening. So your formula gets a $ value that is far above actual. If your organization used activity based accounting you’d have actual hours spent per task and you wouldn’t have to do this…[Read more]

  • Tricia Stuart became a registered member 2 days, 19 hours ago

  • Load More