Card Set Information

2015-05-16 09:12:45
Show Answers:

  1. FINER method
    • 1) feasible
    • 2) interesting
    • 3) novel
    • 4) ethical
    • 5) relevant
  2. positive controls
    those that ensure a change in the dependent variable when it is expected
  3. Negative controls
    ensure no change in the dependent variable when no change is expected
  4. confounding variables
    error that results when a causal variable is associated with two other variables but is not accounted for; may falsely say two variables are associated
  5. observational studies
    can be cohort: those in which subjects are sorted into two groups based on differences in risk factors and then assessed at various intervals

    cross-sectional studies: attempt to categorize patients into different groups at a single point in time

    case-control studies: start by identifying the number of subjects with or without a prticular outcome
  6. Hill's criteria
    describe the components of an observed relationship that increase the likelihood of causality in the relationship

    1) temporality: exposure must occur before the outcome (independent before dependent)

    2) strength: as more variability in outcome is explained by variability in study variable, the realtionship is more likely to be causal

    3) dose-response relationship:as the study or independent variable increases, there is a proportional increase in the response; the more consistent it is, the more likely it is to be causal 

    4) consistency: the relationship is found ot be similar in mutiple settings

    5) plausibility: reasonable mechanism for the independent variable to impact the dependent variable supported by literature

    6) consideration of alternate explanations: remaining explanation is more likely 

    7) experiment; if experiment can be performed, a causal relationship can be determined

    8) specificity: change in outcome varibale is only produced by an associated change in independent variable

    9) coherence: new data and hypothesis are consistent with the current state of scientific knowledge
  7. detection bias
    results from educated professionals usign their knowledge in an inconsistent way
  8. selection bias
    subjects are not representative of the target population
  9. observation bias
    hawthorne effect: people chang ebehavior because they are being studied
  10. internal validity
    suppor for causality
  11. external validity
    generalizability: ability to apply findings to another population

    low generalizability: sample not reflect target populaton

    high: samples represent target population
  12. statisticalyl significant
    not result of random chace
  13. clinical significance
    practical importance of a treatment effect - whether it has a real genuine, palpable, noticeable effect on daily life.
  14. internal validity
    ability to infer causality from a study or replicate its results under the same conditions
  15. skewed distribution
    one that has a tail on one side of the data site
  16. negatively skewed distribution
    mean is lower than median and tail is to the left
  17. positively skewed distribution
    mean is higher than median adn tail is to the right
  18. interquartile range
    • first quartile: multiply n by 1/4
    • --> if a whole number: the quartile is the mean of the value at this position and the next highest position; if it is a decimal, round up to the next whole number and take that as the quartile position

    • third quartile: multply n by 3/4
    • --> follow same instructions

  19. outlier
    any values that fall more than 1.5 IQRs below first or above third quartile
  20. Standard deviation
    • 1) first, determine value of mean
    • 2) find diffrerence between each data point and mean and then square the value
    • 3) add the squared values and divide by n-1
    • 4) take square root of that
  21. null hypothesis
    always a hypothesis of equilance
  22. alternative hypothesis
    nondirectional or directional
  23. t-tests
    rely on the standard distribution or the closely related t-distribution

    1) a test statistic is calculated and compared to a table to determine the likelihood that that statistic was obtained by random chance: called our p-value
  24. What do we do with our p-value
    compare it to a significance level (alpha)

    • p > α: we fail to reject null hypothesis, meaning there is not a significant differenc ebetween the two populations
    • p < α: we reject the null hypothesis and state that there is a statistically significant difference between the two groups

    if the alternative hypothesis is not directional, compare our p-value to α/2
  25. type I error
    value of α is the level of risk that we are willing to accept for incorrectly rejecting the null hypothesis

    likelihood that we report a difference between two populations when one does not exist
  26. type II error
    occurs when we incorrectly fail to reject the null hypothesis

    likelihood that we report no difference between two populations when one actually exists

    sympbolized by Beta
  27. confidence intervals
    determine range of values from the sample mean and standard deviation

    we begin with a desired confidence level (ex: 95%) and use a table to find corresponding z and t score; when we multiply from teh mean, we create a range of values and can then say we are 95% confident that the sample falls between those two values