Short Answer Flashcards
Short Answer (22 cards)
Describe the Relationship between a sample, population, statistic, and a parameter.
A population is the entire group of individuals that you are interested in studying. a sample is a group selected from the population to participate in the research study. A parameter is a characteristic, usually a numerical value, of a population. A statistic is a characteristic of a sample.
What are the basic characteristics of an experiment that differentiate this method from other types of research?
The goal of an experiment is to establish a cause-and-effect relationship b/w 2 variables. Using 2 features - the IV and DV and the control of an experiment.
Why do we need a Null Hypothesis?
We cannot prove the experimental hypothesis using statistics, but we can reject the null hypothesis.
What are the three tricky issues in relation to ethics?
- informed consent
- deception
- social value
How is ethics determined and what is it based off of?
It is culturally determined and they are based on the Nuremberg Code and Ethical guidelines from groups such as the APS/APA and NHMRC
What does a measure of variability indicate?
The degree to which individual observations are clustered about or, equivalently, deviate from that average value.
What are the percentages corresponding to the standard deviations?
1 standard deviations = 68%
2 standard deviations = 95%
3 standard deviations = 99.7%
What are the 2 broad threats to validity?
Internal and external.
Name 6 (out of 17) internal threats to validity.
- confounding
- 3rd variable
- prevent inferring causal relationship
- selection bias
- time of day
- history
- maturation
- Repeated testing
- instrument change
- regression to the mean
- mortality/attrition
- experimenter bias
- selection interaction
- small sample size
- small power
- ordering effects
- multiple tests of significance
Name 4 (out of 7) external threats to validity.
- aptitude
- treatment
- interaction
- situation
- pre-/post-test effects
- reactivity (placebo, novelty, & Hawthorne - being studied)
- Rosenthal effects (self-fulfilling prophecy/Pygmalion effect)
Can a test be reliable and invalid?
Yes. The instrument could be consistently wrong but the consistency means it is reliable but the test is not valid.
What does a t-test do?
It evaluates the difference in means between the groups. It also determines whether a set of scores are from the same population.
What does an independent t-test test for?
(group vs placebo) There are 2 experimental conditions & different participants for both (between-groups), parametric, normal, homo of var.
What does a Dependent t-test test for?
(before & after) With 2 experimental conditions & same participants for both (within-group).
What does an ANOVA test for?
3+ means that are the same :. test H0 that all group means are equal.
Explain why power is important.
- probability that the test will reject the H0 when the H0 is false ie. not make a Type II Error.
- as power increases, chances of a Type II Error decreases.
- Calculate minimum sample size required so that one can be reasonably likely to detect an effect of a given size.
Explain and Identify a Type I Error.
(aka false positive) α
Incorrectly rejecting the H0 - result of something really unlikely just happening to occur in the study.
eg. True (H0) = patient healthy; False (H1) = not healthy –> reject H0 :. falsely accept H1 = sick when healthy ie. Dijana’s blood issue.
“Failing to believe the truth”
“convicting an innocent person”
Explain and Identify a Type II Error.
(aka false negative) β
Incorrectly accepting/failing-to-reject the H0 - results from insufficient power to detect the effects.
eg. True (H1) = patient not healthy; False (H0) = healthy –> reject H1 :. falsely accepting H0 = healthy when sick
“believing the falsehood”
“letting a guilty person go free”
What are the three criteria for evaluating research?
- Construct validity
- Internal Validity
- External validity
List three examples of constructs
self-esteem, attitude, intelligence, or agression.
What happens if there is low construct ability?
The results say little regarding the investigated variables.
In regards to internal validity, do 2 related variables mean a cause-effect relationship?
No.
