Inferential Statistical Tests Flashcards
The selection of an inferential statistical test is based on several factors: The first factor is the ____ of ____ of the ____ to be ____. (When a study has independent and dependent variables, it is the measurement scale of the ____ ____ that is of interest since it is the data on the ____ ____ that will be ____ with the ____ ____.) The second factor is the ____ of the ____. Characteristics of the research design that determine which statistical test to use include the ____ of __ or, if there is only one IV, the ____ of its ____; whether the groups are ____ or ____; whether there are any ____ ____ that need to be ____; and the number of __.
Scale of Measurement; Data to be Analyzed; Dependent Variable; Dependent Variable; Analyzed; Statistical Test; Design of the Study; Number of IVs; Number of its Levels; Independent or Correlated; Extraneous Variables; Controlled; DVs
____ ____ ____ ____: Inferential statistical tests are classified as ____ or ____. Use of both types of tests is based on the assumption that the sample has been ____ ____ from the ____. In addition, these tests assume that ____ are ____, which means that a subject’s participation in the study or performance on the DV is not affected by or related to the ____ or ____ of ____ ____ ____ (except, of course, to the extent that performance is due to the ____ of the __).
Parametric Versus Nonparametric Tests; Parametric or Nonparametric; Randomly Selected; Population; Observations are Independent; Participation or Performance of Any Other Subject; Effects of the IV
Parametric and nonparametric tests differ in several important ways. ____ ____, which include the t-test and analysis of variance, are used to evaluate hypotheses about population means, variances, or other parameters. These tests are appropriate only when the variable of interest has been measured on an ____ or ____ ____ and when certain assumptions about the ____ ____ are met: The first assumption of is that the value of interest is ____ ____ in the ____. The second assumption is that, when a study includes more than one group, there is ____ that is, the variances of the populations that the different groups represent are equal. Violation of these assumptions — especially the assumption of ____ can increase the probability of making ____ ¬_ or ____ _ ¬____.
Parametric Tests; Interval or Ratio Scale; Population Distribution(s); Normally Distributed; Population; Homoscedasticity; Homoscedasticity; Type I or Type II Error
The parametric tests are relatively “____” regarding ____ of their ____. which means that some deviation from a normal curve or from ____ will not necessarily ____ the ____ ____. The most effective way to maximize the robustness of a parametric test is to have an ____ ____ of ____ in ____ ____. Robustness is also increased by having a ____ ____ ____ and setting ____ at a ____ ____ (e.g., at .01 rather than .05).
Robust; Violation; Assumptions; Normal Curve; Homoscedasticity; Invalidate the Test’s Results; Equal Number of Subjects in Each Group; Large Sample Size; Alpha at a Lower Level
The parametric tests are not as robust with regard to the assumption of ____ of ____, however: Even a small amount of dependence among observations can increase the probability of making a ____ _ ____ above the probability indicated by ____.
Independence of Observations; Type I Error; Alpha
____ ____ are used to analyze data collected on variables that have been measured on a nominal or ordinal scale. These tests do not make the same assumptions about the ____ of the ____ ____, as the parametric tests and are referred to as ____-____ ____.
Nonparametric Tests; Shape or the Population Distribution(s); Distribution-Free Tests
The nonparametric tests are used to evaluate hypotheses about the ____ of a ____ rather than the distribution’s mean, variance, or other parameter. Included in this category are the ____-____ ____, the ____-____ _ ____, and the ____ ____-____ ____, shortcoming of the nonparametric tests is that, because they involve using less precise (nominal or ordinal) data, they are less ____. In other words, a researcher is less likely to detect a ____ ________ with a nonparametric test than with a parametric one. For this reason, parametric tests are ____ ____ whenever they can be ____ used.
Shape of a Distribution; Chi-Square Tests; Mann-Whitney U Test; Wilcoxon Matched-Pairs Test; Powerful; False Null Hypothesis; Always Preferred; Legitimately
When picking an inferential statistical test, the first thing you want to consider is the ____ of ____ of the ____ to be ____. (When the study has independent and dependent variables, it is the data on the __ that will be ____.) This will help you distinguish between the tests for ____, ____, and ____/____ ____. The next consideration is the ____ of the ____ ____ - i.e., the number of __ (or, for the chi-square test, the total number of ____); if one IV, the ____ of ____, the nature of the ____ (independent or correlated); the presence of ____ ____; and the number of __.
Scale of Measurement of the Data; DV; Analyzed; Nominal, Ordinal, and Interval/Ratio Data; Nature of the Research Design; IVs; Variables; Number of Levels; Groups; Extraneous Variables
____ ____ and ____ of ____: Regardless of which inferential statistical test is used, its results yield a “____ ____” (_, _ etc.) that allows the investigator to determine whether the obtained sample value is in the ____ or ____ ____ of the ____ ____. This is done by comparing the ____ ____ to a ____ ____, which is the number that corresponds to the ____ that ____ the ____ ____ into ____ and ____ ____. Critical value tables are usually included in the appendices of statistics textbooks. For most statistical tests, when the ____ ____ ____ or ____ the ____ ____, this means that the sample value lies in the ____ ____; when the statistic is ____ ____ the ____ ____, the sample value is in the ____ ____.
Critical Values and Degrees of Freedom; “Test Statistic” (t, F); Rejection or Retention region; Sampling Distribution; Test Statistic to a Critical Value; Boundary that Divides the Sampling Distribution; Rejection and Retention Regions; Test Statistic Equals or Exceeds; Critical Value; Rejection Region; Less Than the Critical Value; Retention Region
The magnitude of the critical value is determined by two factors: ____ and the ____ of ____. Alpha determines what proportion of the sampling distribution represents the ____ ____ and the degrees of freedom (df) determine the distribution’s exact ____. ____ of ____ are the number of values or categories in a distribution that are “free to vary” given that certain values or categories are known or fixed. To illustrate, assume that a sample mean is being used to estimate a population mean, and that the sample consists of eight test scores. The degrees of freedom in this example are 7 (8 - 1) since, once the sample mean is known, only 7 of the 8 scores are “free to vary”: If the sample mean is 50 and seven of the test scores are 80, 80, 70, 50, 40, 40, and 30, the 8th score is not free to vary. It must be 10 in order for the sample mean to be equal to 50.
Alpha; Degrees of Freedom; Rejection Region; Shape; Degrees of Freedom
The method used to calculate degrees of freedom depends on the ____ ____. For the t-test for a single sample, the sampling distribution is based on the ____ ____; and the degrees of freedom are derived from the ____ ____ of ____ (_ - ). When using a single-sample chi-square test, a different kind of sampling distribution is used. It is based on the number of ____ (____) of the ____ and the degrees of freedom are derived from the ____ ____ of ____ ( - _).
Statistical Test; Sample Size; Total Number of Subjects (N - 1); Categories (Levels); Variable; Total Number of Categories (C – 1)
The inferential statistical tests are categorized as parametric or nonparametric. Both types of tests share the assumptions that the sample has been (1) ____ from the population and that observations are (2) ____. A (3) ____ test is used when the data being analyzed represent a(n) (4) ____ or ratio scale of measurement and when two assumptions have been met. The first assumption is that, in the population from which the sample was drawn, scores on the variable are (5) ____ distributed. The second assumption is that there is “homoscedasticity,” which means that the (6) ____ of the populations that the different groups represent are equal.
(1) randomly selected; (2) independent; (3) parametric; (4) interval; (5) normally; (6) variances
If the data to be analyzed do not meet the assumptions for a parametric test or if they represent a(n) (7) ____ or ordinal scale of measurement, a (8) ____ test is used. The major limitation of the nonparametric tests is that they are less (9) ____ than parametric tests, which means that, when using these tests, it is more difficult to detect a false (10) ____ hypothesis.
(7) nominal; (8) nonparametric; (9) powerful; (10) null
Parametric and nonparametric tests both yield a test statistic that the researcher compares to a critical value, which is the cutoff point that divides the (11) ____ into the regions of likely and unlikely values. The critical value for a particular research study is determined by two factors: (12) ____ and the degrees of freedom.
(11) sampling distribution; (12) alpha
The ____-____ ____ is used to analyze the frequency of observations in each category (level) of a nominal variable (or other variable that is being treated as a nominal variable). For example, the ____-___ ____ would be appropriate for comparing the number of people who prefer one of four political candidates or comparing the number of managerial and non-managerial employees who say that either physiological, safety, social, self-esteem, or self-actualization needs are most important.
Chi-Square Test; Chi-Square Test
The chi-square test is used to determine if the ____ of____ (sample) ____ is ____ to the ____ of ____ ____. The expected frequencies are predicted by the ____ ____ and reflect either ____ ____ among the ____ or the known ____ in the ____. Use of the chi-square test not only requires the data to be reported in terms of ____ but also that the expected frequency for any one category is no less than ____ and that observations are ____ (i.e., each subject can appear in only ____ ____).
Distribution of Observed Frequencies is Equivalent to the Distribution of Expected Frequencies; Null Hypothesis; No Differences; Categories; Frequencies; Population; Frequencies; Five; Independent; One Category
There are two forms of the chi-square test – the ____-____ and the ____-____. To help remember when each test is used, it’s better to think of them as the “____-____” and “____-____” ____-____ ____. The former is used when a study includes only ____ ____; the latter is used when the study includes ____ or ____ ____. (When counting the number of variables for the chi-square test, you do not distinguish between ____ and ____ ____.)
Single-Sample and the Multiple-Sample; “Single-Variable” and “Multiple-Variable” Chi-Square Tests; One Variable; Two or More Variables; Independent and Dependent Variables
The ____-____ ____-____ ____ (also known as the goodness-of-fit test) is used when a descriptive study includes only one variable and the data to be analyzed are the frequency of observations in each category of that variable.
Single-Sample Chi-Square Test
Example: The psychologist in Study #2 (treatment of Schizophrenia) is interested in finding out if the 120 patients in her study differ with regard to family background in terms of She determines how many patients have either (1) one biological parent who has received a diagnosis of Schizophrenia; (2) both biological parents who have received a diagnosis of Schizophrenia; or (3) both biological parents who have not received a diagnosis of Schizophrenia. Because this study is a descriptive study that includes only one nominal variable, the appropriate statistical test is the ____-____ ____-____ ____ and the degrees of freedom are equal to: (C — 1) = (3 — 1) = 2.
Single-Sample Chi-Square Test
The ____-____ ____-____ ____ (also known as the chi-square test for contingency tables) is used when a descriptive or experimental study includes two or more variables and the data to be analyzed are the frequency of observations in each category.
Multiple-Sample Chi-Square Test
Example: The psychologist in Study #2 wants to determine if a subtype of Schizophrenia is related to family background and determines each patient’s subtype and family history. The expanded study can be considered either a descriptive study with two variables or an experimental study with one independent and one dependent variable (family background is the IV and subtype is the DV). Because the study includes more than one variable and the analysis will involve comparing frequencies in each category, the appropriate test is the ____-____ ____-____ ____, and the degrees of freedom are: (C - I)(R - 1) = (3 - 1)(5 - 1) = 2(4) = 8.
Multiple-Sample Chi-Square Test
Statistical tests for ordinal data include the ___________________________.
Mann-Whitney U Test, the Wilcoxon Matched-Pairs Test, and the Kruskal-Wallis Test
Example: The psychologist in Study #2 assesses the effects of a phenothiazine on cognitive functioning by comparing the Full Scale IQ scores of patients who have received a moderate dose of the drug to the scores of patients who have received a placebo. The 10 scores turn out to be very skewed. Because the scores violate the assumption of normality for parametric tests, the psychologist converts the IQ scores to ranks and uses the ____-____ _ ____ to analyze the data.
Mann-Whitney U Test
____ ____-____ ____-____ ____: The ____ ____ used when a study includes two correlated (matched) groups and the difference scores on the DV for pairs of subjects are converted to ranks.
Wilcoxon Matched-Pairs Signed-Ranks Test; Wilcoxon Test
Example: The psychologist in Study #2 assigns patients to groups by matching them on the basis of their scores on a measure of premorbid adjustment and randomly assigning members of each matched pair to either the phenothiazine group or the placebo group. Because the study involves correlated groups and the IQ scores are not normally distributed, the psychologist calculates a difference score for each pair of matched patients converts the difference scores to ranks, and uses the ____ ____ to analyze the data.
Wilcoxon Test
The ____-____ ____ is similar to the Mann-Whitney U test but can be used when a study includes two or more independent groups and the data to be analyzed are ranks.
Kruskal-Wallis Test
Example: The psychologist in Study #2 expands her study by including four groups — one that will receive a high dose of the phenothiazine, one that will receive a moderate dose, one that will receive a low dose, and one that will be given a placebo. If the IQ scores violate one or both assumptions for a parametric test, the psychologist can convert the IQ scores to ranks and use the ____-____ ____ to analyze the data.
Kruskal-Wallis Test
Study Tip: The main thing you want to remember about the tests for ordinal data is that they can be described as “____ ____” to specific parametric tests. The ___-____ is the nonparametric alternative to the t-test for independent samples; the ____ is the nonparametric alternative to the t-test for correlated samples; and the ____-____ is the nonparametric alternative to the one-way ANOVA.
Nonparametric Alternatives; Mann-Whitney; Wilcoxon; Kruskal-Wallis
The chi-square test is used to analyze data collected on a nominal variable (or other variable that is being treated as a nominal variable). The (1) ____ sample chi-square test is used when a study includes only one (2) ____ and data on that variable are reported in terms of (3) ____. It would be the appropriate statistical test when a researcher asks a sample of 40 students to indicate which of five textbooks they prefer for an introductory clinical psychology class. The degrees of freedom for this study are equal to (4) ____.
(1) single-; (2) variable; (3) frequencies; (4) (5 - 1) = 4
The (5) _____ sample chi-square test is used when a study includes (6) ____ or more variables and the data to be analyzed are frequencies in each nominal category. It would be the appropriate statistical test when a hospital director wants to compare the number of current adult male, adult female, adolescent male, and adolescent female in-patients who have received a primary diagnosis of either Major Depressive Disorder, Bipolar I Disorder, Bipolar Il Disorder, or Mood Disorder NOS. The degrees of freedom for this study are equal to (7) __________.
(5) multiple-; (6) two; (7) (4 - 1)(4 - 1) = 9
The Mann-Whitney U test and the Wilcoxon matched-pairs test are used to analyze data when a study includes one (8) ____ variable with two levels and the data to be analyzed are reported in terms of (9) ____. The Mann-Whitney U test is appropriate when the two groups are (10) ______, while the Wilcoxon test is used when the groups are (11) ______.
(8) independent; (9) ranks; (10) independent (unrelated); (11) correlated (related)