1. no statistical difference is detected between treatment groups or p value is >0.05. Which Hypothesis do you accept?
    Ho is the null hypothesis – The null hypothesis is accepted when no statistical difference is detected between treatment groups. Typically in clinical studies, researchers will study a medication versus placebo or standard of care. In biostatistics, the null hypothesis is accepted when the p-value is greater than 0.05.
  2. A study was done and they found a statistical significance EXISTS between the groups or p-value was <0.05. Which hypothesis will you accept and which one will you reject?
    Ha is the alternative hypothesis – This is the opposite of the null hypothesis; when Ho is rejected, Ha is accepted. In drug related studies, generally this means that there is a significant difference between the treatment and placebo (or standard of care) groups.When the p-value is low, the null hypothesis (Ho) must go!
  3. Independent and Dependent Variables. While 

    Dependent variables are variables that will DEPEND upon the independent variables. What does INDEPENDENT variables depend on?
  4. Independent
    variables are variables that are set by the researcher. These variables are the ones that we (researchers) can control. 

    • The independent variables are what we set as researchers. The independent variables are Ibuprofen
    • use and Naproxen use. The dependent variable is what DEPENDS on the independent variables. We do
    • not control the BIMS score, so that will be the dependent variable.
  5. Assuming no statistical difference is detected between a new drug and placebo or p value is >0.05. Which Hypothesis do you accept?
  6. The null hypothesis (Ho) would be that
    there is no difference between the two groups. The alternative hypothesis (Ha) would be that there is a difference in blood pressure.
  7. In their study the researchers found that madeuprolol significantly lowers blood pressure more than placebo with a p-value of 0.01. What happens to:
    (1) Null Hypothesis?
    (2) Alternate Hypothesis?
    (3) What type of error stands to be made here?
    • So we would reject the null hypothesis in this case. In regards to Type 1 error, 0.01
    • would equal a 1% chance that a Type 1 error has been committed. Or stated another way: there is a 1%
    • chance that madeuprolol does NOT reduce blood pressure more than placebo based upon the results of
    • this study.

    • If later on, the results of this study are proven false, a Type 1 error would have been committed.
    • By convention, a <5% chance (or a p value of less than 0.05) is considered statistically significant.
  8. How do you define a Type 1 error?
    Type 1 error is defined as detecting a difference when in reality one doesn’t actually exist.

    • In regards to Type 1 error, 0.01
    • would equal a 1% chance that a Type 1 error has been committed. Or stated another way: there is a 1%
    • chance that madeuprolol does NOT reduce blood pressure more than placebo based upon the results of
    • this study.
  9. The formular for power is Power = 1 – Beta

    Whats the benchmark for Power in studies?
  10. In studies, the benchmark Power is generally desired is 80%, which would be equivalent to a beta value
    of 0.2.
  11. Propbablity of making a TYPE 1 error can be checked with which value?
  12. Probability of making a TYPE 2 error can be checked with which value?
    • Beta value
    • If power is 80% than beta is 20%. Thereby the probability of a TYPE 2 error is 20%.

    Type 2 error is the percentage or chance of NOT detecting a difference when one actually exists.
  13. Confidence interval and Statistical Significance.

    95% confidence interval for the change in hemoglobin is (0.3 - 1.5)

    what would make this study statistically significant or not?
    • If this range contained the value 0, we
    • would deem that the 95% confidence interval did not reveal statistical significance
  14. In statistical variables, there are Categorical and Quantitative.
    Whats the difference between categorical and quantitative?
    What are the two types of categorical variables?
    (1) A categorical variable cannot have any value between two points while Continuous variables have a distinct, measurable distance between each value

    • Nominal-Name, yes or no
    • Ordinal-order, pain scale, surveys
  15. what is Parametric Data Versus Non-Parametric Data?
    • Parametric-Continous data
    • Non Parametric-Categorical (ordinal vs nominal)
  16. Discrete variables are continous variables but not all continous variables are discreet. Whats the difference between the two?
  17. Discrete variables are a countable type of variable, but how this differs from continuous, is that you
    • can’t have fractions of a number. A couple examples is number of Accuchecks done per week or CHF
    • exacerbations per year. There is a number associated, but you can’t have a fraction (i.e. the data is not
    • continuous) like 1.36 CHF exacerbations or Accuchecks. The discrete variable will be a whole number
    What type of tests to use for:
    Nominal data
    Ordinal data
    Nominal Data: Chi-squared

    Ordinal Data: Mann-Whitney U-Test and Wilcoxon Signed-Ranks Test
  19. what are other references associated with continuous data?
    • normal distribution
    • parametric tests
    • discrete variables
    • quantitative variables
  20. Appropriate test for continuous variable (ratio or interval) is?
  21. The appropriate statistical test for
    • continuous (ratio or interval) variables will be the t-test or ANOVA(analysis of variance). The t-test (sometimes referred to as student’s t-test) will be used for 1 independent variable with two or less
    • groups. The ANOVA will be used for multiple groups (usually 3 or more, but can be used for 2).
  22. Differential the different types of tests?
    Image Upload 1
  23. WHICH literature is generally peer-reviewed and contains original research done by scientists
    and researchers. A clinical trial done by a pharmaceutical company would be a classic example
  24. literature is compiling primary and secondary literature and creating a larger, broader
    view of clinical information. A classic example of this would be a textbook on pharmacokinetics
  25. literature consists of interpretations and evaluations of primary literature. A
    review article breaking down a bunch of clinical trials relating to a given topic would be an example of
    secondary literature.
Card Set
BCPS Eric Staistics