1/145
Looks like no tags are added yet.
Science
A field of inquiry that relies solely on objective data to determine what is true.
Validity
The requirement that a measurement accurately assesses the specific variable it intends to measure.
Reliability
Consistency or repeatability in measurement; it is necessary for validity but not enough on its own to guarantee it.
Hypothesis
A testable prediction based on quantifiable data, often derived from a theory.
Sample: A subset of a population used in research because studying the entire population is usually impossible.
Population
The entire group (e.g., all people with a certain disease) that a researcher wants to draw conclusions about.
Inferential Statistics
Objective methods used to determine the probability that sample results (like a difference in means) occurred by chance.
Experimental Research
A study where the researcher manipulates at least one variable (IV) and uses random assignment to test for cause and effect.
Non-Experimental Research
Research where the researcher does not control the IV, making it impossible to test causal hypotheses.
Independent Variable (IV)
The variable manipulated by the researcher (the "cause"); in non-experimental research, this is called a subject variable.
Dependent Variable (DV)
The outcome variable that "depends" on the independent variable (the "effect").
Directionality Problem
The difficulty in non-experimental research of knowing which variable is the cause and which is the effect.
Third Variable Problem
The possibility that an unmeasured outside factor is responsible for the relationship between two variables.
Interaction Effect
When the effect of one independent variable on the dependent variable changes depending on the level of another independent variable.
Nominal Level
Data consisting of discrete, mutually exclusive categories or labels with no numeric value (e.g., gender, religion).
Dichotomous Variable
A specific type of nominal variable that has only two possible categories.
Ordinal Level
Data representing a position in a hierarchical ranking where the exact distance between ranks is unknown.
Interval-Ratio Level
Quantities measured on a uniform scale where distances between values are equal and all math operations can be performed.
Continuous Data
Another term for interval-ratio data, implying there is no clean break between values (e.g., weight, GPA).
Frequency Distribution
A graphical array showing how often each value in a data set occurs.
Mean
The average value calculated by adding all scores and dividing by N; it is highly sensitive to outliers.
Median
The middle value in a data set; it is superior to the mean when outliers or skew are present.
Mode
The most frequently occurring value in a data set; rarely used as a primary measure for interval-ratio data.
Outlier
An extremely high or low score that can pull the mean away from the center of the distribution.
Positive Skew
A distribution where the mean is larger than the median due to extreme high scores (tail points to the right).
Negative Skew
A distribution where the mean is smaller than the median due to extreme low scores (tail points to the left).
Standard Deviation
Conceptually, the average distance of all scores in a data set from the mean.
Variance
The average squared distance from the mean; it is the standard deviation squared.
Null Hypothesis (H0)
The assumption that there is no real effect and any sample difference is due to chance or sampling error.
Alpha Level (α)
The standard (usually .05) used to decide if the probability of chance is low enough to reject the null hypothesis.
Type I Error
A "false alarm" where a researcher rejects the null hypothesis even though it is actually true.
type 2 error
Accepting null hypothesis when you should have rejected it
Z-test
A test used to compare a sample mean to a population mean when the population standard deviation is known.
1-Sample t-test
A test used to compare a sample mean to a population mean when the population standard deviation is unknown.
Independent Samples t-test
A test used to compare the means of two separate, independent groups.
Signal-to-Noise Ratio
The logic of the t and F statistics, where the "signal" is the effect of the IV and "noise" is the error variance.
1-Way ANOVA
An Analysis of Variance used to compare more than two means while preventing the buildup of chance error.
F-ratio
The statistic for ANOVA, calculated as Mean Square between (signal) divided by Mean Square within (noise).
Pearson's r
A correlation coefficient measuring the degree to which two interval-ratio variables correspond with each other.
Multiple Regression
uses multiple predictor variables to predict a single outcome (criterion).
Coefficient of Determination (R2)
The percentage of variability in the outcome variable that can be explained by the predictor(s).
Chi-Square (χ2)
An inferential test used for nominal data to see if observed frequencies differ from what is expected by chance.
The Normal Curve (50/34/14 Rule)
In a normal distribution, 50% of scores fall below the mean, and approximately 34.13% fall between the mean and a Z-score of +1.0.
Decomposing Z-scores
To find the area for a non-whole Z-score (like 1.25) in Table A, split the value: the row is determined by the tenths (1.2) and the column by the hundredths (.05).
Standard Error of the Mean (SEM)
The standard deviation of a hypothetical distribution created by taking an infinite number of sample means of size N from a population.
Cohen's d
The specific statistic used to measure effect size for a Z-test, calculated as the difference between the sample and population means divided by the population standard deviation (d=(M−μ)/σ).
Effect Size Standards (Cohen's d)
Standards for magnitude are: .2 = small, .5 = moderate, and .8 = large.
Platykurtic Distributions
Smaller sample sizes make the sampling distribution of the mean flatter (platykurtic), requiring higher critical values from the t-table than the Z-table.
Degrees of Freedom (1-Sample t)
Calculated as N−1, representing the number of values in a data set that are free to vary.
The "Noise" Factor
In Z or t formulas, the denominator represents "noise" (variability); as the standard deviation increases, the denominator grows, making it harder to reject the Null Hypothesis.
Sample Size Impact
Larger sample sizes decrease the size of the denominator, which increases the final Z or t value and makes rejecting the Null Hypothesis more likely.
Homogeneity of Variance
An assumption for the independent samples t-test stating that the standard deviations of the two groups being compared should not be significantly different.
Familywise Error
The problem where running multiple t-tests instead of one ANOVA increases the probability of a Type I error (e.g., with 6 comparisons at α=.05, the error risk jumps to 26%).
Tukey's HSD
A post-hoc test used only after a significant ANOVA to find exactly which mean differences are large enough to be "real".
HSD Calculation
Calculated by multiplying a value from Table F by the square root of (Mean Square within / n).
Interaction Precedence
A critical rule stating that interaction effects take precedence over main effects when explaining the results of a study.
Interaction Explanation
An interaction is explained by describing how the effect of one variable "depends on" or is "opposite for" the levels of another variable.
Standard Error of the Difference (SEDiff)
In a paired-samples t-test, this denominator is adjusted by subtracting a term (2⋅r⋅SEM1⋅SEM2) that accounts for the correlation between scores.
The Regression Plane
In multiple regression with two predictors, the relationship is visualized as a three-dimensional plane rather than a two-dimensional line.
Coefficient of Determination (R2) in multiple regression
In multiple regression, this value represents the percentage of variability in the criterion (Y) that can be predicted by the entire set of predictors.
Coefficient of Contingency (C)
The specific effect size statistic used for Chi-Square tests.
Chi-Square Effect Standards (C)
Standards for magnitude are: .10 = small, .25 = moderate, and .40 = large.
null hypothesis for z test
sample mean = population mean
Null hypothesis for 1-sample t-test
𝜇=𝜇0
Null hypothesis for independent samples t-test
μ1= μ2
Null hypothesis for chi-square test
fo=fe
Null hypothesis for one-way ANOVA
µ1=µ2=µ3....
null hypothesis for factorial ANOVA
µ1=µ2=µ3.... for each sample
Null hypothesis for Perason's R
p=0
if the correlation between X and Y is zero, the best predictor of the value of Y is the mean of X.
T/F
false
In a linear regression analysis, the higher the Pearson's r, the lower the values of the residuals
T/F
true
In addition to an interval-ratio level predictor,a nominal predictor can also be used in linear regression analysis.
T/F
true
In linear regression analysis, if the slope is +.50, which of the following is true?
A. you can predict 25% of the variance in Y from the value of X.
B. you can predict 50% of the variance in Y from the value of X.
C. for every unit decrease in X there will be a 1/2 unit increase in Y.
D. None of the above are true.
D. none of the above
In linear regression, if the slope is -.25, which of the following is true?
A. Every unit increase in X will result in a 1/4 unit increase in Y
B. Every unit increase in X will result in a 4 unit increase in Y
C. Every unit increase in X will result in a 1/4 unit decrease in Y
D. None of the above are true
C. Every unit increase in X will result in a 1/4 unit decrease in Y
In linear regression, the change in value of Y that occurs with each unit change in X is referred to as the
A. slope
B. coefficient of determination
C. residual
D. intercep
A. slope
In linear regression, the intercept is the value of X when the value of Y is 0.
True
False
false
In linear regression, the mean value of all residuals will be greater for a correlation of -.70 than for a correlation of +.60.
True
False
false
In linear regression analysis, the intercept is determined by what factors?
A. the mean of Y
B. the mean of X
C. the slope of the regression line
D. all of the above
D. all of the above
The regression line represents the line that
A. minimizes the residuals when using the value of X to predict the value of Y
B. intersects the Y axis at the point where the X variable equals 0
C. slopes downward when the correlation between X and Y is negative
D. all of the above
D. all of the above
The slope of the regression line is determined by
A. the sign of Pearson's r for X and Y
B. The absolute value of Pearson's r for X and Y
C. The standard deviation of both X and Y
D. All of the above
D. All of the above
When all of the points in a scatterplot lie on the regression line, the correlation between the two variables must be
A. positive
B. statistically significant
C. + 1.0 or - 1.0
D. Over .50, but less than + or - 1.0
C. + 1.0 or - 1.0
In a case where two predictors are both significantly correlated with the criterion, but where the multiple R is NOT significant, what should one do?
go back and check your calculations because you must be wrong
Just use the mean of the criterion to predict the value of Y for any two predictor values
go ahead and develop the multiple regression plane equation anyway
perform simple linear regression with the one predictor that has the highest Pearson r with the criterion
perform simple linear regression with the one predictor that has the highest Pearson r with the criterion
The larger the multiple R the ___________ from __________ on average will be the predicted values of Y (the criterion)
farther from; the mean of Y
closer to: the mean of Y
farther from; 0
closer to; 0
farther from; the mean of Y
The lower the multiple R, the ___________ the ____________ will be each point in the 3-dimensional scatter plot.
closer to; regression line
farther from; regression line
closer to: regression plane
farther from; regression plane
farther from; regression plane
As the correlation between the two predictor variables increases, what will happen to the value of multiple R?
It will increase
It will decrease
It will remain unchanged
You cannot tell without the actual values of all three Pearson rs
It will decrease
When both the independent and dependent variables are interval/ratio level, which statistical test should be used?
A. paired-samples t
B. independent samples t
C. Pearson's r
D. either A or C, because the same people (or items) are being measured twice.
C. Pearson's r
Why is the term 2r(SEM1)(SEM2) not included in the independent samples t-test?
A. Because we only use SEM1 in the independent samples t-test
B. Because the relationship in an independent samples t-test is assumed to be non-linear
C. Because the r is assumed to be 0 in the independent samples t-test
D. Because the denominator in the independent samples t-test underestimates the estimated standard error the difference
C. Because the r is assumed to be 0 in the independent samples t-test
The term -2r(SEM1)(SEM2) in the paired samples t-test causes the t-ratio to be ________ than it would if the term were excluded.
A. larger
B. smaller
C. more biased with respect to the population parameter
D. less biased with respect to the population parameter
A. larger
With respect to the paired-samples t-test, as the size of the r increases, the t-ratio _____ making it ________ to reject the null hypothesis.
A. increases; less likely
B. increases; more likely
C. decreases; less likely
D. remains the same; equally likely
B. increases; more likely
If the Pearson's r between variables X and Y is -.50, how much of variance in Y can be predicted by X?
A. 50%
B. 25%
C. none - negative correlations do not permit you to explain variance in a criterion variable
D. you cannot tell from the Pearson's r how much variance in Y can be predicted from X.
B. 25%
Which of the following, if any, is true with respect to Pearson's r?
A. Either of the variables must be ordinal
B. The ranges of values for X and Y can be no greater than + or minus 1 SD of their means.
C. The form of the relationship between X and Y can be non-linear
D. None of the above are true
D. None of the above are true
Which of the following is used in determining the degrees of freedom between in the 1 way ANOVA?
A. the number of independent variables
B. the number of subjects or scores
C. the number of levels of the IV
d. none of the above
C. the number of levels of the IV
In the F statistic, the sum of squares between is influenced by
A. variation within each group in the design
B. differences between each group mean and the grand mean
C. the number of groups (or means being compared) in the study
D. A and B
E. B and C
E. B and C
In an experiment, the numerator in the F statistic is a measure of
A. the effect of the independent variable on the dependent variable
B. the effect of individual differences among subjects on the dependent variable
C. the effect of the dependent variable on the independent variable
D. the effect of the number of subjects on the dependent variable
A. the effect of the independent variable on the dependent variable
In a one-way ANOVA, as the sum of squares within increases, the probability of accepting the null hypothesis
A. decreases
B. increases
C. remains constant
D. you cannot tell because the degrees of freedom within changes depending on the value of the sum of squares within.
B. increases
If the F statistic is significant in the 1-way ANOVA, the next step to take would be
A. to report the result and reject the null hypothesis.
B. to report the result and accept the null hypothesis
C. to accept the null hypothesis and conduct a post-hoc test
D. to reject the null hypothesis and conduct a post-hoc test
D. to reject the null hypothesis and conduct a post-hoc test
Why do we conduct a post-hoc test if we find a significant F in a one-way ANOVA?
A. Because the book and Prof. Kilianski told us to.
B. Because the sum of squares within underestimates the population sum of squares
C. Because you don't know which means differ from which other ones when there are more than two of them.
D. Because smaller sample sizes are less likely to yield significant differences, and each group is a smaller sample than the total N.
C. Because you don't know which means differ from which other ones when there are more than two of them.
Assuming that the mean differences and standard deviations are constant, you are less likely to reject the null hypothesis when using a paired-samples t-test than an independent samples t-test.
True
False
false
In a 1-way ANOVA, if the sum of squares between remains constant and the number of means being compared increases, you are more likely to find a significant difference among the means.
True
False
false
In a 1-way ANOVA, both the number of levels of the independent variable and the size of the sample influence the degrees of freedom within.
True
False
true