PSY 230 Final Exam
If the power of an experiment is .7, the probability of making a Type II error is ______.
.3
The sum of all the individual deviation scores will always equal ____.
0
Power can vary from:
0 to 1
Beta equals: (equation)
1 - power
If you retain the null hypothesis, you may be making (two options):
1. A correct decision 2. Type II error
A bar chart should NOT be used for (3):
1. Continuous variables 2. Ratio level variables 3. Interval scales of measurement
Use a bar graph with data of what kind of measurement scales? (2)
1. Nominal (categories) 2. Ordinal (rank orderings rather than raw scores)
The analytic technique that is best for assessing the effects of two independent variables on a dependent variable is known as:
2-way ANOVA
What type of ANOVA is required to analyze the following data: There are 2 factors, 1. Caffeine intake before the exam (0 mg, 200 mg, or 400 mg) 2. Hours of sleep the night before the exam (0-3 hours, 4-6 hours, or 6+ hours)
3x3 factorial ANOVA: two factors, both of which have 3 levels
If SS total = 20 and SS between = 14, SS within = ?
6
Pearson's r is:
A measure of the paired scores
The two-way ANOVA allows what?
An assessment of the interaction between two independent variables
The variability within groups (MS within) is assumed to be due to:
Chance
In a/an _______-groups design, you use the same raters, and in a/an ________-groups design, you use different raters
Correlated; independent
Descriptive statistics vs. Inferential statistics
Descriptive: describe or characterize data (central tendency, variability, shape, etc.) Inferential: use obtained sample data to make inferences and draw conclusions about the population (probability, t-tests, etc.)
If F obtained is less than 1, we:
Do not need to compare it to F critical (the treatment did not have a significant effect)
The null hypothesis for an F test states:
Each group is a random sample from populations where the group means are equal
Omega squared and eta squared are measures of:
Effects size
ANOVA is also known as a/an...
F test
The ________ increases with the effect of the independent variable
F value
True or False: larger samples will always have larger values of the range.
False
True or False: Spearman's rho is the slope of the least squares regression line when the scores are plotted as z scores
False; Pearson r
Fahrenheit or Celsius are measured on the ____ scale of measurement while Kelvin is measured on the ____ scale.
Interval; ratio
When the line is horizontal, there is or is not a main effect? When the line is NOT horizontal, there is or is not a main effect?
Is not Is
If the probability level of an experiment is less than or equal to the alpha level, you ____ the null hypothesis
Reject
If the z-obtained value is in the critical region, what is the appropriate conclusion for the test?
Reject null hypothesis
How is eta squared calculated?
SS between / SS total
When calculating F critical, the DF numerator is:
The DF between (k-1)
When calculating F critical, the DF denominator is:
The DF within (N-k)
The larger F obtained becomes, the more reasonable it is that:
The IV had a real effect
The main effect of factor A is defined as:
The effect of factor A averaged over all of the levels of factor B
"Jumping the gun" (Saying there is an effect when there really isn't) is an example of which error?
Type I error
If a researcher sets a conservative alpha of .001, they are likely concerned about which type of error?
Type I error
"Missing the boat" (Saying there is not an effect when there really is one) is an example of which error?
Type II error
If alpha is changed from .05 to .01, the probability of a Type ___ error increases
Type II error
To ethically increase the probability of rejecting Ho, you should:
Use the most powerful test
The advantage of a powerful experiment is that:
You are more likely to detect the real effects of the IV, if there are any
A type II error occurs when:
You incorrectly retain a false null hypothesis (false negative) Ex: Let a guilty person free in trial
The proportion of variance accounted for by a correlation between 2 variables is determined by ____.
r squared
In a normal distribution, what z-score value separates the lowest 50% of all the scores from the rest of the scores? (What is the mean?)
z = 0.00
Factors that affect power (4) and their relationship (inverse or direct)
1. Alpha level (direct relationship: increase alpha, increase power) 2. Sample size ( direct relationship: increase sample size, increase power) 3. Effect size (direct relationship: larger effect size will have greater power) 4. Sample variability (inverse: the greater the sample variability, the lower the power to detect a real effect)
Characteristics of a sampling distribution of a statistic (3):
1. Gives all the values a statistic can take 2. Gives the probability of getting each value of a statistic under the assumption it is due to chance alone 3. Is a probability distribution
Use a histogram with data of what kind of measurement scale? (2)
1. Interval (equal intervals between adjacent units, NO absolute zero point) 2. Ratio (equal intervals between adjacent units, there IS and absolute zero point)
As the difference between the group means increases, these increase as well (3 things)
1. MS between 2. Treatment effect 3. F obt
Conditions for a binomial distribution (5)
1. There is a series of N trials 2. On each trial, there are only 2 possible outcomes (dichotomous) 3. Outcomes are mutually exclusive 4. Trials are independent 5. The probability of each possible outcome on any trial stays the same for each trial
What is true about an a priori (planned) comparison? (3)
1. They have higher power than post hoc comparisons 2. Used only when there is adequate theoretical support 3. Use the same basic formula as the single-sample t test
Types of posteriori (post hoc) comparisons (2)
1. Tukey's HSD (Honestly Significant Differences) test: most common, controls for all pair-wise mean comparisons 2. Scheffe test: more conservative, controls for every comparison
When interpreting F(2,18), how many participants were involved?
21 F(DF between, DF within), DF within = N - k = 18, k = 3
The two-way ANOVA results in the calculation of ___ F ratios
3
When interpreting F(2,18), how many groups were examined?
3 F(DF between, DF within), DF between = k - 1 = 2
In a normal distribution, what percent of scores fall between a z-score of -1 and the mean?
34.13%
The more non-parallel the lines are in an interaction plot, the _______ the strength of the interaction.
Greater
After conducting a one-way ANOVA, you get a p value of .161. What does this mean in terms of HOV and significant difference?
HOV was not violated (p is greater than .05) There was no significant difference between the means of one or more groups tested (p is greater than .05)
A distribution is negatively skewed when most of the scores occur at the higher or lower end?
Higher
When analyzing the results of ANOVA tests, if your p value is less than or equal to .05, what is violated?
Homogeneity of Variance (HOV)
An alpha level of .05 indicates that:
If the null hypothesis is true, the probability of falsely rejecting it (making a Type I error) is limited to .05
Restricting the range of X or Y will increase or decrease the correlation?
Increase
Researchers want the statistical power of the test to be relatively high (at least .80) because this will:
Increase the probability that a false null hypothesis will be rejected
In ANOVA tests, the word "factor" denotes the:
Independent variable
When the effect of one factor is not the same at all levels of another factor, a/an _____________ is said to have occured
Interaction
When the lines are parallel in the interaction plot for a 2-way ANOVA, there is or is not an interaction?
Is not (if lines are not parallel, there is probably an interaction)
You're exploring the differences in exam scores (DV) associated with varying amounts of sleep (IV). You categorize the amount of sleep into groups: high, medium, and low. In terms of ANOVA, these groups are called:
Levels
The one-way ANOVA partitions total variability into two groups:
MS between and MS within (MS = variance estimate)
Maximizing power in an experiment ________ beta
Minimizes
What is the most desirable amount of power for an experiment? The most common?
Most desirable: .80 Most common: .40-.60
Why don't we use multiple t tests instead of ANOVA to compare group means?
Multiple t tests would increase (inflate) the probability of a Type I error
Spearman's rho should be used when examining the relationship between two variables when one or both are measured on the __________ scale
Ordinal
Using letter grades (A, B, C, D, F) to classify student performance on an exam is an example of measurement on a/an ______ scale of measurement
Ordinal
When analyzing the results of ANOVA tests, one or more group means are significantly different if:
P value is less than or equal to .05
In an experiment with 8 subjects involving a non-directional alternative hypothesis, the obtained results were 7 pluses and 1 minus. To evaluate the null hypothesis, what probabilities would you add together?
P(0) + P(1) + P(7) + P(8)
The reason we don't use multiple t-tests when comparing more than two groups is due to:
Preserving the alpha level
If F obtained is equal to or greater than F critical, we ________ the null hypothesis
Reject
An example of a test that is designed to compare all possible pairs of means while maintaining the Type I error is a:
Scheffe test
What constitutes a small effect size when calculating omega squared and eta squared? Medium? Large?
Small: .01-.05 Medium: .06-.13 Large: .14 and beyond
The closer the points on a scatter plot are to the regression line, the ____ the correlation is between the scores.
Stronger
The variability between groups (MS between) is assumed to be due to:
The effect of the independent variable
A main effect for variable A means that:
The effect of variable A is significant when averaged over all levels of variable B
A significant interaction effect indicates that:
The effects of one variable are NOT the same at all levels of the other factor
The two-way ANOVA assesses:
The effects of two independent variables in one experiment
The between groups estimate increases with the effect of:
The independent variable
The effect of one factor averaged over the effects of another factor is called:
The main effect
The larger F obtained becomes:
The more reasonable it is that the independent variable had a real effect
The Pearson r coefficient is used to determine what, and of what units?
The relationship between 2 variables which are measured on different units
In a repeated measures design experiment...
The same subjects receive both treatments
What does a 2x3 factorial ANOVA mean? Give an example
There are 2 factors (IVs), and the first factor has 2 levels and the second has 3 levels Ex: Factor 1: sex (male or female), Factor 2: Treatment (1, 2, or 3)
A type I error occurs when:
You incorrectly reject a true null hypothesis (false positive) Ex: convict an innocent person
When a variable is continuous, the probability of event A occuring equals:
the area under the curve corresponding to A divided by the total area under the curve
The null hypothesis in the one-way ANOVA asserts that:
μ1 = μ2 = μ3 = μN