Card Set Information

2013-03-19 14:27:27

Show Answers:

  1. All students in 230 n=400, In the population 60% are woman, four professors, two sections.

    Students are sorted by gender and 60 females and 40 males?
    Quota Sampling
  2. A random number generator on a calculator is used to select 100 enrolled students?
    Random Sampling
  3. Each of the 400 students are included?
    Census Sampling
  4. Two of the four instructors are selected, then one of the four sections is selected. Within each section 25 students are randomly selected?
    Multistage sampling
  5. Compute the Standard Error of mean?
    Divide the SQR of n by σx .

    • n=49;σx=5      =5/√49
    • n=100;σx=5    =5/√100
  6. Population mean?
  7. Population Standard Deviation?
  8. Sample Mean?
  9. Sample Standard Deviation?
  10. Mean of distribution of Sample Means?
  11. Standard Error of the mean?
  12. What happens to the Standard Error or the mean(σm) as n increases?
    It gets smaller
  13. What happens to the Standard error of the mean(σm) as σx, increases?
    It gets higher/larger
  14. Central Limits Theorem?
    • Principle 1: μm=μx
    • Principle 1:the mean of the distribution of sample(μm) means equals the population mean(μx)

    Principle 2:σm=PSR of σx^2/n= σx/PSR of n, where n is the sample size.

    Principle 3: The distribution of the sample means will be normal if either of the two conditions are met : 1.) The population of individual scores is normally distributed  or the sample size is equal to or exceeds 30.
  15. What should a research modify if she wants to reduce the standard error of the mean?
    Reduce her sample size, n
  16. As the desired degree of confidence increases, the margin of error__________.
  17. As the same size increases the margin of error_________?
  18. For the Z-test the LOS is .05 Critical value is +-?
  19. For the Z-test when the LOS is .01 the critical value is +-___?
  20. When the null hypothesis is retained, is the alternative hypothesis retained or rejected?
  21. When the null hypothesis is retained the alternative hypothesis retained or rejected?
  22. When the null is retained, do we claim the intervention as Significant or Not Sig?
    Not sig.
  23. When the null is rejected, do we claim it is sign or not sig?
    Statistically Significant
  24. Rejecting a true null?
    Type I Error.(alpha)
  25. Rejecting a false null?
    Correct rejection of Ho(Power)
  26. Retaining a true null?
    Correct retention of Ho(1-alpha)
  27. Retaining a false null?
    Type II error.(Beta)
  28. Alpha at .05 and power equals .70?
    Probability of retaining a true null?
    Retaining a true null(1-alpha)=(1-.05)=.95
  29. alpha at .05 and power =.70?
    Probability for rejecting a true null?
    Rejecting a true null(alpha)=(.05)=.05
  30. Alpha at .05 and Power at .70? 
    Probability for retaining a false null?
    Retaining a false null(Beta)=(1-Power)=(1-.70)=.30
  31. Alpha at .05 and Power at .70?
    Probability for rejecting a false null?
    • Rejecting a false null(1-beta)
    • (Beta= Power-1=.30)=(1-.30)=.70
  32. Suppose you change your mind about alpha .05, and change it to .01? How does this affect you?
    If you lower alpha then it will be too difficult to reject the null.
  33. Alpha .01, Power at .70? Probability of retaining a true null?
    1-alpha =(1-.01)=.99
  34. What is the probability of rejecting a true null?
    If alpha is .01 and Power is .70.
  35. Probability that you will retain a false null gone down, up, or the same?
    (1-power=Beta) Beta has gone up.
  36. Probability that you will reject a false null has gone up, down, same?
    (1-beta) has gone down.
  37. Trueblood wants to avoid a Type 1 error so he decides to set alpha at .001. What is the problem with doing this?
    When you lower alpha to .001 you increase room for making a Type II error, because the alternative hypothesis will have more odds and you will lower your power.
  38. Power increases as the effect size______ as n ________ and the the LOS ________.
    • Effect size increases;
    • N increases;
    • LOS increases.
  39. Larger effect sizes can be obtained by _______ the population of standard deviation (σx) and by _______the strength of intervention.
  40. If a researcher has a low power and reject HO, is the effect size small or large?
  41. If a researcher has a high power and retains HO is the effect size small or large?
  42. Whereas with the Z test the population is _________.
  43. With the T test the population standard deviation is?
  44. The population standard deviation must be estimated with the___ test.
  45. Whereas for the Z test, as N increases, the critical value for rejecting the null hypothesis _______.
    Remains the same.
  46. For the T test as N increases the critical value for rejecting the null _______.
  47. If n=11 LOS=.05 will a research have less, the same or more power if she is able to use the z test as opposed to the T test. Why?
    More power because with a z-test they only have to work with a population size being effected by two factors. (LOS and directionality) With a T test they have a smaller sample size and have 3 factors affecting their small pool, leaving room for false occurrences.