GU Research Mod 9

The flashcards below were created by user MeganM on FreezingBlue Flashcards.

  1. Inferential statistics are based on the laws of ________.
  2. Inferential statistics provide a means for _________ about a population, given data from a sample.
    drawing conclusions
  3. The tendency for statistics to fluctuate from one sample to another.
    sampling error
  4. A theoretical (rather than actual) distribution b/c no one draws consecutive samples from a population and plots their means.
    Sampling distribution
  5. The SD of a sampling distribution of the mean.
    Standard error of the mean (SEM)
  6. The word ______ signifies that the various means in the sampling distribution have some error as estimates of the population mean.
  7. The smaller the SEM, the _______ the means are as estimates of the population value.
    • more accurate 
    • (less variable)
  8. ___________ is used to estimate a parameter - for example, a mean, a proportion, or a mean difference b/w two groups.
    Parameter estimation
  9. With interval estimation, researchers construct a _________ around the estimate.
    Confidence interval (CI)
  10. What are confidence limits?
    Upper and lower limits of confidence interval
  11. A ________ distribution is the probability distribution of the number of "successes" in a sequence of independent yes/no trials, each of which yields "success" with a specified probability.
  12. What is the best way to get representative samples?
    probability samples
  13. To estimate population parameters, it is advisable to use _______ samples.
  14. Researchers use inferential statistics to estimate _______ ________________ from _________ statistics.
    • population parameters
    • sample
  15. Inferential statistics provide a framework for making _______ judgments about the _______ of sample estimates.
    • objective
    • reliability
  16. The ________ of statistical calculations depends on the extent to which results from a sample are similar to what you would have obtained had you randomly selected people from the population.
  17. _________ involves calculating a single descriptive statistic to estimate the population parameter.
    point estimation
  18. _____ estimates convey NO info about margin of error so inferences about the accuracy of the parameter estimate cannot objectively be made.
  19. ________ estimation is useful because it indicates a RANGE of values within which the parameter has a specified probability of lying.
  20. With _____ estimation, researchers construct a confidence interval (CI) around the estimate.
  21. Constructing a CI around a sample mean establishes a range of values for the population value as well as the probability of what?
    being right! (the estimate is made with a certain degree of confidence)
  22. THe CI is rarely ________ around a sample proportion.
  23. The ______ of the CI depends on the value of the proportion AND the sample size.
  24. The larger the sample size, the _______ the CI.
  25. The CI for a proportion never extends below ____ or above _____.
    • 0
    • 1.0
  26. CI convey importan info about the _______ of the estimate.
  27. Hypothesis testing is based on _______ inference.
  28. Researchers make a ______ error by rejecting a null hypothesis that is actually true.
    Type I
  29. False negative conclusion.
    Type II error
  30. A ____ error might allow an ineffective drug to come onto the market.
    Type I
  31. A ________ error might prevent an effective drug from coming onto the market.
    Type II
  32. Researchers control the risk of a _______ error by selecting a level of significance.
    Type I
  33. A level of significance signifies the probability of ___________ a true null hypothesis.
    incorrectly projecting
  34. The two most frequently used significance levels are ___ and _____ and are referred to as ________
    • .05, .01
    • alpha
  35. Increasing sample size will reduce the risk of a ______ error.
    Type II
  36. By selecting a significance level, researchers establish a ________.
    decision rule
  37. The decision rule is to reject the null hypothesis if the test statistic falls at or beyond the limits that establish a _________ on an applicable theoretical distribution and to accept the null hypothesis otherwise.
    critical region
  38. The ________ indicates whether the null hypothesis is IMPROBABLE given the results.
    critical region
  39. In one-tailed tests, the critical region of improbable values is in only one tail of the distribution - the tail corresponding to what?
    the direction of the hypothesis
  40. Easier or harder to reject null hypothesis with a one-tailed test?
  41. Two classes of statistical tests.
    Parametric and nonparametric
  42. _______ tests involve estimation of a parameter, require measurements on at least an interval scale, and involve several assumptions like variables are normally distributed.
  43. ________ tests do NOT estimate parameters and involve less restrictive assumptions about the shape of the distribution.
  44. _______ tests are more powerful than ________ tests.
    parametric; nonparametric
  45. When comparisons involve different people (men vs women), the study uses a ______ design, and the statistical test is a test for _________.
    • between-subjects
    • independent groups
  46. In ________ designs, comparisons are not independent b/c the same people are used in all conditions, and the statistical tests are tests for ________.
    • within-subjects;
    • dependent groups
  47. ____________ refers to the number of observations free to vary about a parameter.
    degrees of freedom
  48. If the absolute value of the test statistic is ________ than the tabled value, the results are statistically significant.
  49. Any computed probability greater than ______ indicates non significance.
  50. A _______ can be used when there are two independent groups and when the sample is dependent.
  51. When there are multiple dependent variables, the risk of a _______ error increases.
    Type I
  52. Researchers sometimes apply a __________ when they run multiple tests to establish a more conservative alpha level.
    Bonferroni correction
  53. The Bonferroni correction tends to increase the risk of ________ error.
    Type II.
  54. Incorrectly concluding there is no statistical association when in fact there is one.
    Type II error
  55. ______________ can be constructed around the difference b/w two means, and the results provide info about BOTH statistical sig and precision of estimated difference.
    Confidence intervals
  56. ________ is richer and more useful in clinical applications than p values.
  57. A _______ is computed from pretest and posttest data.
  58. The ______ test involves assigning ranks to the two groups of scores.
    Mann-Whitney U
  59. The _________ test involves taking the difference b/w paired scores and ranking the absolute difference.
    Wilcoxon signed-rank
  60. ________ is the parametric procedure for testing differences b/w means when there are three or more groups.
    Analysis of Variance (ANOVA)
  61. The statistic computed in ANOVA.
  62. ANOVA decomposes total _________ in a ______ variable.
    • variability
    • dependent
  63. Variation between groups is contrasted with variation within groups to get _______.
    an F-ratio
  64. When differences b/w groups are large relative to variation within groups, the probability is _______ that the independent variable is related to or has caused group differences.
  65. The ___________is the sum of sum of squared deviations of individual group means from the overal grand mean for all participants.
    sum of squares between groups
  66. The ______ is the sum of the squared diviations of each individual score from its OWN group mean.
    sum of squares within groups
  67. The function of __________ is to isolate the differences b/w group means that are responsible for rejecting the overall ANOVA null hypothesis.
    multiple comparison procedures (post hoc tests)
  68. In multiple comparison procedures, why is it NOT appropriate to use a series of t-tests?
    b/c this would increse the risk of a Type I error
  69. __________ is used to test the relationship b/w one categorical independent variable and a continuous dependent variable.
    One-way ANOVA
  70. __________ concerns whether the effect of one independent variable is consistent for all levels of a second independent variable.
  71. ________ ANOVA is used when thre are 3 or more measures of the SAME dependent variable for each participant.
  72. Sum of squares-treatments is the same as the sum of squares-between in regular ANOVA and represents the effect of the ________ variable.
  73. The sum of squares-error is similar to the sum of squares-within in regular ANOVA and both represent variations associated with _______.
    random fluctuations
  74. The sum of squares-_____ has no counterpart in a simple ANOVA. Why?
    • subjects
    • b/c those being compared are not the same people
  75. The SS-subjects term captures _______.
    individual differences
  76. _____ ANOVA yields a more sensitive test of the relationship b/w the IV and DV than ________ ANOVA.
    • RM
    • Between-subjects
  77. What is removed from the denominator in computing the F-statistic?
    variability attributable to individual differences
  78. The ______ test is a generalized version of teh Mann-Whitney U test.
  79. The __________ test is based on assigning ranks to the scores of various groups.
    Kruskal-Wallis or Mann-Whitney U
  80. The ________ test is used when the number of groups is greater than two and a one-way test for indepenent samples is desired.
    Kruskal-Wallis or Mann-Whitney U
  81. The ________ test is for ANOVA by ranks .
  82. The _________ test is used to test hypotheses about group differences in proportions, as when a contingency table has been created.
  83. The _______ test enables us to decide whether a difference in propertions of a certain magnitude is likely to reflect a real tx effect or only chance fluctuations.
  84. _________ are the cell frequencies that would be found if there was no relationship b/w the two variables.
    Expected frequencies
  85. The _______ statistic is computed by summarizing differences b/w observed and expected frequencies for each cell.
  86. When is a chi-square test not appropriate?
    • when sample size is small (30 or less), OR
    • when there are cells with small frequencies (5 or fewer)
  87. ___________ test should be used to test the significance of differences in proportions.
    Fisher's exact
  88. When the proportions being compared (pretest-posttest design is used to compare changes in proportions on a dichotomous variable) are from two paired groups the appropriate test is ________.
  89. What is the correlation coefficient calculated when two variables are measured on at least the interval scale?
    Pearson's r
  90. Pearson's r summarizes the _______ and ________ of a relationship b/w two variables.
    • magnitude
    • direction
  91. Pearson's r can be used both _____-group and ______-group situations.
    • within
    • between
  92. If test scores are obtained before and after an event, the correlation b/w the two scores is a ______-group situation.
  93. Pearson's r is a ________statistic.
  94. When the assumptions for a parametric test are violated, or when the data are ordinal level, the appropriate coefficient of correlation is ____________.
    • Spearman's rho OR
    • Kendall's tau
  95. The chi-square statistic is computed by comparing ______ frequencies with ______ frequencies.
    • observed
    • expected
Card Set:
GU Research Mod 9
2015-11-11 00:43:51
Research polit beck
Mod 9 Ch 17 p. 404-422
Show Answers: