GoConqr has a new look! Let us know what you think.

Flashcards

Week 8

Indah Epidemiology English 14 cards

Cards in this deck (14)

  1. 1Examples of Cohort studies in book - name, year, association?British Doctors Study; 1951, Smoking & Lung Cancer. US Nurses' Health Study; 1976, Oral Contraceptive & Breast Cancer. Avon Longitudinal Study of Parents & Children (ALSPAC), 1990, indi. genes combined with enviro pressures to influence health & dev'ment over life course
  2. 2Measures of association in Cohort and Experimental studiesAbsolute Risk, RR, Risk difference (AR), AR%, Odd Ratio
  3. 3Absolute riskOnly for prospective studies. No comparison group. Number of ppl developing disease / number of ppl followed
  4. 4Relative riskRisk in exposed / risk in unexposed. 1 = no association. More than 1 positive association. Less than 1 negative association
  5. 5Risk difference / Attributable riskAmount of disease risk that can be attributed to a specific exposure. Risk exposed - risk unexposed
  6. 6AR%AR expressed as proportion of the risk in exposed. Risk difference / Risk in exposed
  7. 7RR vs Attributable riskRR measures the strength of the association. AR measures how much of the disease risk is attributable to a certain exposure
  8. 8Odds ratioodds for cases / odds for controls
  9. 9Statistical inferencedrawing conclusions about a population from a sample
  10. 10Random sampleequal chance of selection. no criteria or preference
  11. 11Hypothesis Testing - Null Hypo & Alternate HypoRisk for exposed same risk for unexposed (Null Hypo, RR/OR = 1). Risk for exposed NOT same as risk for unexposed (Alt. Hypo RR/OR not equal to 1).
  12. 1295% confidence interval95% confience that lower and upper limits of this interval include the true mean, proportion, rate or risk for the entire population where its been sampled.
  13. 13Confidence IntervalShows the range within which the true value for the population is likely to lie. if it include 1 - no statistically significant diff b/w exposed and unexposed.
  14. 14p-valueif value is more than 0.05, confidence interval will include 1 diff may be due to chance & no real diff = not statistically sig. If less than 0.05, Conf.I doesnt include 1, result due to chance is small diff observed is true - statistically significant.

Information

Description
Epidemiology Flashcards on Week 8, created by Indah on 16/10/2013.
Created by
Indah
Views
17
Rating

Feedback

What is it about?

What do you think of the new look?

How do you like it?