Research Methods Flashcards
-Observational Studies -Experiments -Hypothesis Testing -Correlation Research
The goals of psychology are to describe, explain, predict and control behavior and mental processes. To attain these goals psychologists engage in two types of research:
Basic research and applied research
What is basic research?
Basic research is concerned with gathering knowledge for the sake of gaining knowledge
What is applied research?
Applied research is conducted for a specific purpose such as problem solving and improvement of quality of life. It uses basic research and APPLIES it to specific problems.
What are the most frequently used research methods psychologists use to collect data?
-naturalistic observation
-the case study
-the survey
-correlational study
-experimental study
What is a naturalistic observation?
Observing people in their natural environment, without interacting with them
What is a case study?
Studying one person or situation in depth
What is a survey?
Asking the same question of each member of a sample group of people
What is an experimental study?
A controlled study where the researcher directly changed factors to intersect whether there’s a change
What are the 4 essential components in all experiments?
Independent variable
Dependent variable
Experimental groups
Control group
What changes in an experiment and what doesn’t change?
Changes: Dependent Variable
Doesn’t Change: Independent Variable
What is known as the treatment in the experiment?
The independent variable. In an experiment, the researcher manipulates the independent variable in order to determine if it caused any change.
The experimental group is given the __________, which is the change in the ____________ variable.
treatment; independent
What is random sampling? What does “random” mean?
Random sampling helps obtain a representative sample from a list of all the members of the population you’re interested in. “Random” means that every member of the population has an equal chance of being included in the sample.
__________ _____ occurs when participant are assigned to groups in such a way rust certain characteristics are more likely to be selected.
Selection bias
A census undercounts certain people. Name them.
The poor
Undocumented immigrants
People without permanent addresses
What is a census?
A census tries to contact EVERY individual
What is under-coverage?
Some parts of the population is left out when you chose the sample.
What is a response bias? Give examples.
A general term used to describe a number of different conditions or factors that cue respondents to provide inaccurate or false answers during surveys or interviews.
-question wording
-interviewer characteristics
-questionnaire distribution
-untruthful answers
-interviewer induced bias (an interviewer influences response)
-non-response bias
What is non-response bias?
A form of under-covering. Your bias is because of the missing people in your sample.
How can double negatives lead to inaccurate results?
Double negatives can lead to wording bias misunderstanding. “Dos is seem possible or impossible…”
Poorly worded question can lead to inaccurate results.
______ __________ indicate the expected answer or the respondent.
Leading questions
What is social desirability?
Social desirability is a type of interviewer-induced response bias. People tend to give answers they think the interviewer wants to hear, or that they think will make them look better in the eyes of the interviewer.
What is false consensus bias? What can expose this bias? Give an example of consensus bias.
People often have a tendency to overestimate how much other agree with us. This is called false consensus bias, and one reason why observational studies are important is that they can expose this bias.
Ex: “If I feel this way and I’m a sane person, then other sane people much feel the same way.”
Social Desirability vs False Consensus
-you want to ask the interviewer for a date
-John tells the priest-interviewer that he goes to church
Social desirability
Social Desirability vs False Consensus
-Who wouldn’t love a bulldog?
-Of course most people will vote for my candidate
False Consensus
What is an experiment?
A method of collecting data in which the researcher exerts control over the subject or units in the study. Some treatment is applied to a group and their responses are measured.
What is an experimental design?
Involves choosing treatments and describing how subjects or units are to be assigned to treatments.
Describe an observational study.
-observe without trying to influence
-observes some variable of interest
-doesn’t try to influence the variable
Describe an experiment.
-impose some treatment on a group
-observe the response
-applies some treatment to a group to determine its effect
-to establish cause and effect between 2 variables
What do observational studies and experiments have in common?
It’s important that the group observed or experimented on is representive of the population.
What are experimental units?
The thing a treatment is applied to in an experiment. These can be human or no human.
If experimental units are human, they’re called ________.
Subjects
What is an explanatory variable?
The explanatory variable tris to explain or bring about change in the response variable.
Explanatory variables are also called ________.
Factors
How many factors are there?
In an experiment they tested 2 medicines.
2 factors
Suppose a study wanted to determine the most effective dose of aspirin. These various doses of the explanatory variable, aspirin, are referred to as ______ of the factor.
Levels
Repeating experiments on different subjects or under different conditions is called _______.
What does this help with?
replication
This helps with the accuracy of an experiment.
What is generality?
To apply finding to the larger population our treatment groups represent.
What is the Hawthorne effect?
People change their behaviors simply because they know they are being watched.
Why do people try to decrease the Hawthorne effect?
It is necessary to make sure that two groups (treatment and control) are alike.
What are lurking variables?
Variables you aren’t aware of or taking into account, but might be causing a response.
When we can’t tell which variable might be unwind an effect we say the variable is __________.
confounded
True or false.
Lurking variables can be confounding variables.
True
What is a way to limit confusion due to lurking/confounding variables?
Have two groups:
Control group
Treatment group
What conditions are ideal when you have a control group and a treatment group?
We want:
-the groups to be alike
-the double-blind administration of a treatment and a placebo
What is control in an experiment?
To account for lurking variables
What is the basic idea behind control in an experiment?
Possible confounding variables should operate equally on both groups. Then we can be confident that any difference between the two groups’ responses is die to the treatment, and not some lurking variable we haven’t thought of.
How do you control bias?
By randomizing subjects.
What are the three principles of experimental design?
- Replication
- Control
- Randomization
What is a quasi-experiment?
You are following all experimental procedures, except your sample isn’t random.
In what cases can quasi-experiments be performed and convenient?
They are useful for preliminary studies because they’re often much cheaper and easier to do experiments on random sample. They can also be useful if you’re only interested in getting results for your sample and you don’t need to generalize results to the whole population.
True or False.
You can use a quasi-experiment to prove something conclusively.
False. You can conduct a quasi-experiment, however to prove anything conclusively you’ll need a follow-up study with a random sample.
Hypotheses come in pairs: a _____ hypothesis and an ___________ hypothesis.
null; alternative
What is an alternative hypothesis?
The alternative hypothesis says there’s IS a difference, change, or new finding.
Alternative vs Null Hypothesis
-Vitamins make animals move faster
-Homework causes liver cancer
-Apple eaters have a higher IQ
Alternative Hypothesis
Alternative vs Null Hypothesis
-Vitamins don’t affect speed of snails
-Homework does not cause liver cancer
-Apple eaters have same IQ
Null Hypothesis
__________ _______ is the term Fo the whole process of creating hypotheses and then testing them.
Hypothesis testing
____________ _______ is a part of hypothesis testing—it’s the part where you use laws of probability to decide if your null is likely enough that can’t reject it.
Significance testing
What is a significance level expressed as?
A significance level is expressed as a probability. This is the probability that your results came from random chance alone.
.05 (5%) level of significance vs .01 (1%) level of significance
-not due to chance 95 out of 100
-due to chance 5 out 100
.05 (5%) level of significance
.05 (5%) level of significance vs .01 (1%) level of significance
-The most strict of the two levels
-Not due to chance 99 out of 100
.01 (1%) level of significance
True or false.
If there is a 5% chance, or less that your result came form random chance, reject the null.
True :)
People often have a tendency to fall into the “I knew it all along” syndrome when looking at past events. Sometimes when a person is correct it’s really the result of a lucky guess, not because they had an ability to know all the factors and predict the outcome.
This common effect is called _________ _____.
Hindsight bias
Why is hypothesis testing so good at revealing hindsight bias?
Hypothesis testing is especially good at exposing hindsight bias because it looks directly at the likelihood that the person could have gotten the correct answer due to random chance.
In a ________ ___________, both are going the same direction. In a ________ ___________, they are going opposite directions.
positive correlation; negative correlation
_________ correlation has nothing to do with how weak or strong the correlation is, it has to do with an _______ relationship; when one does up, the other goes down, or, when on does down, the other goes up.
Negative; inverse
What is the correlation coefficient?
The measure of the strength and direction of a correlation is the correlation coefficient. (Check it up for examples)
What is the range of the correlation coefficient? Give an example.
The number ranges from +1 (very strong positive correlation) to -1 (very strong negative correlation). In the middle is 0, which is no correlation at all.
The absolute value in a coefficient indicated relative strength.
Ex: the correlation of -0.45 is stronger than that a correlation of +0.30. The plus or minus sign in the correlation coefficient indicated whether the correlation is positive or negative.
True or False.
Correlation studies prove cause and effect (aka causation)
FALSE. Correlation studies are useful in making predictions. They are also useful for showing that one variable may be a factor in affecting another variable. However, correlation studies DO NOT prove cause and effect.
What is the only way to determine causation?
An experiment
The APA code of ethics states that all research participants must be given a full explanation of the nature and purpose of any study after its conclusion. This process is called __________.
debriefing
A scatter plot in which all the points fall in a straight line with a slope greater than zero would be characteristic of two sets of variables which
are strong correlated.
would show equal amounts of variability.
have a correlation coefficient of 0.5.
none of the above
are strong correlated
Which of the following represents a negative correlation?
There is no relationship between school discipline and student behavior.
Schools that employ strict discipline produce students with poor behavior.
Schools that employ strict discipline produce students with good behavior.
Schools that employ strict discipline produce some students with good and some students with bad behavior.
Schools that employ strict discipline produce students with poor behavior.
A major requirement for a sample to be used in psychological research is that it be
statistically supportive.
practical.
statistically significant.
done
representative.
representative
The coefficient of correlation is a range of numbers that can be between
0.00 to 1.00
-3 SD to +3 SD
+1.00 to -1.00
-1.00 to 0.00
+1.00 to -1.00
Which of the following statements is true?
none of the above
An experiment is the only way to determine causation.
If you have a correlation, then you know causation.
You can conclude causation based on a correlation.
An experiment is the only way to determine causation.
A response variable…..
is not affected by changes in the explanatory variable
is measured after the treatment
is irrelevant to the experimental results
causes systematic changes in other variables
is manipulated by the researcher
is measured after the treatment
