Lec: 63: Evidence-Appraisal: Statistical inference Flashcards Preview

Fundamentals > Lec: 63: Evidence-Appraisal: Statistical inference > Flashcards

Flashcards in Lec: 63: Evidence-Appraisal: Statistical inference Deck (21):

What 3 factors affect study results?

Bias/confounding Chance Treatment/Therapy/Exposure etc


How are bias/confounding minimized?

Strong study design Randomization Masking/blinding


What is standard error of the mean?

Estimate of the standard deviation of all sample means --Describes the precision of the sample estimate --Based on variability and sample size --Measures “how far off” estimate is likely to be from population mean


How is standard error of the mean/proportion calculated?

Estimated std error of mean: s/sqrt(n)

  • = population standard deviation ÷ square root of sample size

Estimated std error of proportion (p): sqrt(p(1-p)/n)

  • (see image)

A image thumb

What is a confidence interval?

A range of “plausible values” for the true population value


How is confidence interval calculated?

Confidence interval = estimate (plus/minus) critical value x standard error


A image thumb

Standard error depends on...

variability and sample size


Critical value depends on...

sample size and confidence


What does statistical significance show?

Results are unlikely to be caused purely by chance


Define Type I error:

Rejecting the null hypothesis (there is a difference) when the null hypothesis is true (false negative)


Define p-value:

The probability of obtaining the observed test statistic, or one more extreme, if the null hypothesis is true


Define Type II error:

Not rejecting the null hypothesis (no difference) when the null hypothesis is false (false positive)


Define beta:

P(Type II error) Probability of concluding there is no difference when a difference exists


Define power:

Power = 1 - beta Study with good power is less likely to “miss” important differences


What is power dependent on?

  • Type I error rate alpha
  • Effect size (e.g. difference in means or proportions)
  • Variability of outcome measure
  • Sample size

Typically first 3 are fixed and sample size is increased to achieve >80% power


What are the steps in hypothesis testing?

1. State hypotheses 2. Choose significance level 3. Calculate appropriate test statistic and corresponding p-value 4. Decide whether evidence is sufficient to reject the null hypothesis


What is the alternative hypothesis?

The study or research hypothesis, want data to SUPPORT the alternative


What is the null hypothesis?

Converse of study hypothesis, want data to REJECT the null


How are hypotheses described?

One sided Two sided


Define alpha:

Alpha = P(Type I error) The probability of rejecting the null hypothesis when it is true

  • Probability of concluding there is a difference when no difference exist
  • overstating the significance of your findings


What is the test statistic?

Test statistic = (Observed value - Hypothesized value)/Standard error of observed value

Decks in Fundamentals Class (66):