PSY 210: Exam 2
To have construct validity, a measure should have both:
Convergent Validity (correlate with other measures that it should be related to) Discriminant Validity (not correlate with measures that it should not correlate with) (EX: Self esteem survey should have questions based on how they feel about themselves, it should not correlate with their age, sibling count, occupation)
Why is high response rate important in survey research?
It makes the survey participants more diverse and representative of the population
How do correlational studies differ from experiments and quasi-experiments?
It only involves dependent variables and is designed to answer predictive and descriptive research questions
Why is construct validity particularly relevant to survey research?
It studies the relationship between the measure of construct and scores on measures of other constructs
Suppose that a researcher found that as introversion of participants go up, the number of social activities attended per week decreases. This finding represents a _______________
Negative relationship
Open-ended response scale
No pre-determined choices
No correlation
No relationship between high/low levels of one and high/low levels of another
Closed-ended response scale
Nominal, Ordinal, or Interval choices already typed out
In inferential statistics, we look for evidence against the ___
Null hypothesis
A ____ hypothesis predicts no effect, whereas an _____ hypothesis predicts an effect.
Null, alternative
In correlational research, multiples measures are collected from _____ group(s) of subjects to determine of those measures co-vary.
One
The decision to reject the null hypothesis in a statistical test is made when
P= alpha, p< alpha
Whenever a strong relationship exists between variables, prediction of one variable from other variables is
Possible
One-tailed scientific hypothesis
Predicts a directional effect
Two-tailed scientific hypothesis
Predicts an effect in either direction
Scientific/alternative hypothesis
Predicts an effect or relationship in the population
Null hypothesis
Predicts no effect or relationship in the population
Match the correct sampling techniques with the example: Volunteers are selected from students such that the proportion of students in each age range in the sample matches the proportion in the population.
Quota Sample
Split-Half Reliability
Randomly divide items into 2 subsets and examine the correlation in total scores across the 2 subsets (Not as common)
Type I error
Rejected the null hypothesis when it is actually true
Positive correlation
Relationship in which high levels of one variable corresponds to high levels of another
Negative correlation
Relationship in which high levels of one variable corresponds to low levels of another
The II error
Retained the null hypothesis when it is actually false
Variation
SD squared
Match the correct sampling techniques with the example: Individuals are selected at random from a list of phone numbers to be called for a survey
Simple random sampling
What are some typical methods of determining internal consistency?
Split-Half Reliability and Cronbach's Alpha
The _________ measure of variability indicates the average difference between the scores of a distribution and the mean of a distribution.
standard deviation
Acceptable terms for causation:
-Cause -Increase/Decrease -Benefits -Impacts -Enhances/Undermines -Effect/Affect -ImpRoves/Boosts -If>Then type statements
Acceptable terms for correlations:
-Get -Have -Linked -More, less -Tied -Connected -Related -Tend -Leads
What are some suggested ways of improving response rate?
-Presentation -Clear instructions -Contacting participants multiple times -Incentives
Alpha level of an inferential test is typically set at __% or less.
5
Target population
A group of individuals who we want to learn about from a study (EX: All US students)
Probability (random sampling)
A sample is selected directly from the target population
Non-Probability (convenience sampling)
A sample is selected from the accessible population
We cannot "support" a null hypothesis with non-significant results from a single study. Instead, we say that the results of the study fail to reject the null. Why?
A series of studies must be conducted
Which of the following sampling techniques involves choosing individuals from the ovulation such that individuals chosen from a pre-existing group? A. Simple Random Sample B. Cluster Sample C. Haphazard Sample D. Stratified Random Sample
B. Cluster Sample
Which of the following sampling techniques is likely to create the smallest amount of sampling error? A. Cluster Sample B. Simple Random Sample C. Volunteer Sample
B. Simple Random Sample
Why do longer surveys tend to yield more reliable responses than shorter surveys?
Because an unusual response on a short survey can skew the results quite a bit.
Why is reliability of responses also important (in addition to validity)?
Because unreliable surveys reduce the validity of the scores of a study
Which of the following describes a quota sample? A. People who respond to a mailed survey B. All students who sign up for a study C. 60% of the population is left-handed so 60% of the participants chosen are also left-handed
C. 60% of the population is left-handed so 60% of the participants chosen are also left-handed
Which of the following measures indicates the most common score in a distribution? A. Mean B. Median C. Mode D. Range
C. Mode
Cronbach's Alpha
Calculate the relationship between the scores for each pair of items on a survey and then determine the average of these correlations for the survey (It doesn't look at just two groups like split-half, it looks at ALL pairs- most common)
Correlational studies are not well-suited for answering _________ research questions.
Causal
Match the correct sampling techniques with the example: Individuals are selected at random for a study from a list of members of Girl Scout troops to represent the population of girls aged 8-13
Cluster Sample
Which one is not a probability sample? A. Simple random sample B. Stratified random sample C. Cluster Sample D. Haphazard Sample
D. Haphazard Sample
What is the goal of correlational studies?
Examine the strength and direction of relationships between variables
T/F: All sampling error can be removed from a study if the appropriate sampling technique is used.
False
T/F: It is usually possible to test the entire population in a research study.
False
T/F: Volunteer samples are uncommon in psychological research.
False
True/ False: A correlational study showed that men who had one or more servings of tomatoes a week were 60% less likely to get prostate cancer. The means that tomatoes causes reductions in cancer rates.
False
The most common type of convenience sample used by psychologist is the ______ sample.
Hahazard
When there are extreme scores (outliers) in a distribution, mean is not a good representation of the typical score in the distribution. Why?
High scores pull the mean value toward the top or bottom of the distribution range
What does internal consistency of scores indicate?
How similar scores on the different items of a survey are compared to one another.
Criterion-related validity can be tested to determine what?
If a survey can predict behaviors
What are inferential statistics?
If the obtained sample means can be used to infer about the population that we want to learn about
What is the goal of null-hypothesis significance (inferential) testing?
If we can can reject the null-hypothesis
How have researchers typically dealt with social desirability bias in survey responses?
Include a measure of this bias to test a respondents level of social desirability in order to allow a measurement of the bias within the sample
Match the correct sampling techniques with the example: Students are selected at random from lists of students at a university organized by class level such that the proportion of individuals at each class level in the population of students matches the proportion selected for the sample.
Stratified random sample
Standard Deviation
The average difference between the scores and the mean of the distribution
Mean
The calculated average of the scores
What does the probability (p) value obtained for the inferential test indicate?
The chance of obtaining the data when the null hypothesis is true
If we find that an inferential test for an experiment is significant, this means that
The condition means significantly differed, the null hypothesis should be rejected
What does test-retest reliability mean?
The consistency of participants responses over time (Giving test multiple times)
Range
The difference between the highest and lowest scores in a distribution
What is sampling error?
The difference in scores from the population to the sample
What is criterion-related validity?
The extent to which a measure distinguishes participants on the basis of a particular behavioral criterion (EX: SAT is valid to the extent that it distinguishes between students that do well in college versus this who do not)
Median
The middle score
Mode
The most common score
Accessible Population
The portion of the target population that is accessible to the researcher, from which a sample is drawn
Sample
The selected participants from the accessible population
What does variability indicate?
The spread of scores in a distribution
When a variable that is not of interests in a study may explain the results of a study, this is called
The third variable problem
What is central tendency?
The value that represents typical score in a distribution (Mean, Median, Mode)
In a research study the p value obtained for the inferential statistic is .07. This means that
There is a 7% chance of obtaining the data when the null hypothesis is true
Why are non-probability samples used frequently in psychological research?
They are more convenient (easier to get participants)
Why are probability samples less likely to have less sampling error than non-probability samples?
They are more representative of the population
As researchers, we usually set alpha at .05 because
This sets our Type I error rate at 5%
What is the main goal when developing a new survey?
To answer descriptive or predictive research questions
What is the goal in sampling?
To choose individuals for the sample who are representative of the behaviors/attitudes of the population
Why is it necessary to set an Alpha level?
To determine how much Type I Error can occur in a test
T/F: A disadvantage of cluster sampling is that a portion of the population may be missed in the sample, giving the researcher a sample that is not fully representative of the population
True
T/F: All studies, regardless of sampling technique, are subject to sampling error.
True
T/F: In convenience samples, the probability of an individual being chosen from the population cannot be determined, because individuals are not chosen randomly.
True
T/F: Quota samples are similar to stratified random samples without the random selection from subsets of the population.
True
Regardless of the type of observational techniques used to measure behavior, the goal of any correlational study is to examine relationships between _______ or more measures of behavior.
Two
If we reject the null hypothesis when it is really true, we have made a
Type I error
If we retain the null hypothesis when it is really false, we have made a
Type II error
What is survey research?
Uses survey observational techniques to measure behavior (Can answer descriptive, predictive, and causal)
Identify the variables and the direction: Fontaine who runs a gym has created a special exercise class for women that are overweight. He finds that for two days a woman attends the class, she loses 1/2 % body fat.
Variables: body fat, attendance days Direction: Negative
Identify the variables and the direction: Prof Johnson finds that for every diet beverage someone drinks during the day, the likelihood of heart attacks is increased by 4%.
Variables: diet beverage consumption, heart attacks Direction: Positive
Identify the variables and the direction: Margaret notices that the temperature in her office drops several degrees from July to Dec. In careful review of the budget, she discovered that the companys utilities expenses also drop.
Variables: temperature, utility expenses Direction: Positive
Identify the variables and the direction: Over a four week period, a teacher watches elementary school children playing during recess. She counts the number of times students slide down a slide. As the weeks progress, she sees little change in sliding behavior.
Variables: time passed, sliding behavior Direction: Not correlated
Match the correct sampling techniques with the example: Individuals are selected from a participant pool made u of students at a university who volunteer for the study.
Volunteer/Haphazard
The p value compared to the alpha level determines what?
Whether you reject or retain the null hypothesis
What are descriptive statistics?
mathematical summaries of results
A _____ hypothesis predicts a directional effect, whereas a _____ hypothesis predicts an effect in either direction.
one-tailed, two-tailed
What is the main limitation of survey research?
people lie about their behavior