Experimental Psych Unit 2
Semantic Differential Scale
worthless to valuable; can be/use any type of adjectives unlike the Likert scale that uses strongly disagree to strongly agree
Leading Question
wording leads people to one response more than others, weakening construct validity (How awesome am I at teaching?); biased
3 ways you could operationalize sleep quality
1.) Physiologically: EEG; 2.) Self-Report: Survey. 3.) Observational: Sleep-study
3 ways you could operationalize stress
1.) Physiologically: sweat response, brain waves, blood pressure, cortisol levels. 2) Self-Report: Survey. 3.) Observational: Response
2 ways you could operationalize academic motivation
1.) Self-report or observational: Hours spent studying. 2.) Observational: Look at grades
Part; Entire
A sample is to _________________ as a population is to _________________.
Observational
Advantage: Helps us look at human behavior
an anonymous survey asking whether students want the campus mascot to be changed.
An in-person institutional review board (IRB) meeting would probably be required for all of the following studies EXCEPT:
random sampling
Another term for probability sampling is:
Divergent
Another word for discriminant validity is ______________ validity.
Causal Inferences; made
Bivariate association claims' failure to meet the criteria of temporal precedence and internal validity means that _________ cannot be_________.
From an ethical standpoint, in what way is researching prisoners with tuberculosis similar to researching children with ADHD?
Both groups of participants have less autonomy than other types of participants.
Avoiding Plagiarism
Cite Sources (According to Sheldon (2002)," ) and try to paraphrase instead of using direct quotations
Physiological
Disadvantage: Everyone's physiology is different (confounding variables); EEG, MRI
Self-Report
Disadvantage: Positive self-presentation; Advantage: Easy and Cheap
Interrater Reliability
Correlation or Kappa; 2 individuals showing consistent ratings of a behavior they are observing ( r > .7)
statement 1; statement 2
Dr. Li is interested in creating a measurement of religiosity. According to the dictionary, religiosity is "the quality of being religious; piety; devoutness." He creates a measure comprising 10 statements. People respond to each statement using the following scale (1 = strongly disagree, 3 = neither agree nor disagree, 5 = strongly agree). Some of the statements are below. Statement 1: "I believe in a religion." Statement 2: "Part of who I am stems from my religious beliefs." Statement 3: "I believe that religion is unnecessary." Dr. Li gives his new measure to a group of 100 adults in order to test the internal validity of his measure. To demonstrate internal validity, we would expect participants with higher agreement on ___________ to also show higher agreement on ______________ (i.e., there would be a positive relationship between which two items?).
He finds that the measure of gambling is not correlated with a measure of life satisfaction in the same two groups of people.
Dr. Reynolds is a clinical psychologist who specializes in treating pathological gambling. Pathological gambling is defined as being unable to resist impulses to gamble. Bothered by not having a good measure that he can give to clients to determine whether they are suffering from this condition, he creates a new measure of pathological gambling. The measure has 15 questions, and it takes 10 minutes to complete. Dr. Reynolds has decided to test the discriminant validity of his new measure. He has a group of first-time Gamblers Anonymous (GA) attendants complete his measure and finds that they score higher than a group of people who do not attend the group. Which of the following results would provide evidence for discriminant validity?
Validity of measures
Face/Content, Criterion, Convergent, Discriminant
It has a double negative
In his measure of "need for cognition" (the degree to which people like thinking and problem-solving), Dr. Jonason asks his participants to rate their agreement with the following statement: "I have never not enjoyed thinking." What is the problem with this question?
Median
Middle number; used with ordinal and equal interval/ratio data (skewed)
Criterion Validity
Objective - something we can measure; most researchers are interested in this; known-groups paradigm; Relate to key behavior?- how well someone will do in graduate school in comparison with GRE score
Discriminant Validity
Objective; distinguishes from other constructs; shows your measurement isn't measuring something else (If you are measuring anxiety, you want to distinguish between something similar; distinguishing stress and anxiety are similar but separate); mostly negative correlations (r's) - r = -.18
Convergent Validity
Objective; positive correlation; taking a measure and seeing if it relates to other similar measurements (ACT vs. SAT scores
Likert Scale
Strongly disagree to strongly agree; type of question that asks for a degree of agreement
Quantitative Measurement
Take on a numerical form: 1.) Ordinal (Order - rank in class); 2.) Interval (intelligence), have a meaning between the numbers (what you get in a race), doesn't have an absolute zero; 3.) Ratio (someone is twice as tall as someone else, percentage in class) has an absolute zero
skewed to the left
The median is greater than the mean in a distribution that is
P Value
The probability level which forms basis for deciding if results are statistically significant; the more low this is, the less likely to have obtained the results if there was, in fact, no association between your variables (P < .05)
For a third variable to be plausible as the explanation in an established association, what must be true?
The third variable must be related to both of the measured variables in the original association.
Forced-Choice Question
people give their opinion by picking the best of two or more options
What purposive, convenience, quota, and snowball sampling have in common
They result in samples where some people are systematically left out.
How do reverse-worded items address shortcuts?
They slow down readers, making them answer more carefully.
a measure is valid but not reliable
What is not possible regarding measures?
3
What is the mode of the following set of scores? 2, 3, 3, 5, 7
2
What is the standard deviation of the following set of scores? 2, 3, 3, 5, 7
a difference in the height between the bars
When examining an association claim using a bar graph, an association is indicated by which of the following?
A description of the study's hypotheses
You submit a study for approval by the institutional review board (IRB), and they tell you that written informed consent is required. Which of the following can be excluded from your informed consent document?
unusual
a biased sample consists of too many __________ cases.
Conceptual Definition
a researcher's definition of a variable at the theoretical level
If people who give favorable answers also come back to the restaurant
a restaurant owner uses a response card with 4 items in order to evaluate how satisfied customers are with the food, service, ambiance, and overall experience. Which result would show the response card's validity?
Open-ended question
allow respondents to answer any way they like
goal of inferential statistics
allows us to draw conclusions about the population we are interested in (using laws of probability)
Bivariate Correlation
an association that involves exactly two variables
IRB (Institutional Review Board)
any academic research needs to be proposed to their IRB, review for ethical violations and/or procedural errors
Mean
average (mew and M) Balance point; distorted by skew or outliers; used with equal interval/ratio (not skewed) data
Standard deviation
average distance from the mean
Core ethical principles
beneficence, justice, respect for rights/dignity, fidelity/responsibility, integrity
Test-retest Reliability
correlation, compare results, similar scores are reliable ( r > .5)
nonsignificant result
can't rule out the chance that the result came from a population in which the association is 0
Restriction of range
condensed/shortened; situation involving a bivariate correlation, not a full range of possible scores on one of the variables in the association, so the relationship from the sample underestimates the true correlation.
Negative Wording Question
contains negatively phrased statements, making its wording complication or confusing, potentially weakening construct validity (Which of the following is not...)
Causal Criteria
covariance, temporal precedence, internal validity (3rd variable problem)
Statistical Validity
degree to which statistical conclusions we draw from a study are accurate and reasonable
Informed Consent
each person learned about the research project, considers its risks/benefits, and decides whether to participate
Fidelity/Responsibility
establish relationships of trust; accept responsibility for professional behavior (in research, teaching, and clinical practice)
Faking Bad
giving answers on a survey that make one look worse than one really is
Statistical Significance
helps researcher evaluate the probability that the result would have occurred if the "null hypothesis" were true (E.g., a 0 association); results depend on effect size and sample size (the larger these are, more likely they are this)
Construct Validity
indication of how well a variable was measured or manipulated in a study
Internal Reliability
items on your measurement are similar to one another; cronbach's alpha; (r > .7)
.50
large or strong effect size
central tendency
mean, median, mode
.30
medium or moderate effect size
Mode
number that occurs the most; peak of distribution; used with categorical data
goal of descriptive statistics
organize and summarize data
Double Barreled Question
problematic because it asks 2 questions in one, weakening construct validity; not sure what your participant is responding to
Categorical Measurement
quantitative/nominal; differ by quality, not by number
Respect of rights/dignity
recognize that people are autonomous agents; protect people's rights, including privacy, right to give consent for treatment/research, right to have participation treated confidentially; understand that some populations may be less able to give autonomous consent, and take precautions against coercing people
Socially desirable (Faking Good)
responding in a way that makes you look better
types of measurement
self-report, observational, physiological; need all 3 of these to determine the best practice and data
Response Set
shortcut respondents may use to answer items in a long survey, rather than responding to the content of each item; Acquiescence and fence sitting
.10
small or weak effect size
Research ethics involves consideration of
society and science, professional issues, treatment of research participants
Fence Sitting
someone who goes down the middle or is neutral; doesn't strongly agree or disagree with anything; researchers avoid this by giving 4 options
Integrity
strive to be accurate, truthful, and honest in one's role as researcher, teacher, or practitioner
Justice
strive to treat all groups of people fairly, sample research participants from the same populations that will benefit from the research, be aware of biases
the principle of justice calls for a balance between
the kind of people who participate in research; the kind of people who benefit from it
Effect Size
the magnitude of a relationship between two or more variables; R or Cohen's d; larger effect sizes lead to more accurate predictions and are usually more important
Face/Content Validity
subjective; Face (It looks like what you want to measure - "looks like a duck") and Content (Measure contains all parts that your theory says it should contain - "Walks like a duck, quacks like a duck")
Acquiescence
tend to strongly agree with everything or give the highest rating
Reliability of measures
test-retest, interrater, internal
Operational Definition
the specific way in which a concept of interest is measured or manipulated as a variable in a study
what is true regarding moderators?
they can inform external validity
Correlation
to occur or vary together (Covary) systematically, as in the case of 2 variables; any study with all measured variables is this
Beneficence
treat people in ways that benefit them, don't cause them suffering, conduct research that will benefit society, benefits should out-weigh the cause
Plagiarism
type of research misconduct along with data falsification and fabrication
An association between 2 quantitative variables
type of statistic that will usually be a Pearson's "r"
An association between 2 categorical variables
type of statistic will be X2 (Squared) test for independence (Ex: Variable #1: Drinking level; Variable #2: Dementia
An association between a categorical and quantitative variable
type of statistic will usually be a "t-test" (Ex: Measured Variable #1: Legality of marijuana - categorical; Measured Variable #2: Purchase of junk food ($) - ratio)
Moderating Variables
variable that, depending on its level, changes the relationship between two other variables (Ex: Gender and texting vs. mental health)
there will be greater construct validity
what is not true of finding a stronger effect size in an association claim?