Study Guide 2 - Research Methods And Data Analysis Flashcards
Critical Thinking
-raising hard questions is part of critical thinking
-critical thinking is the process of checking claims, and making judgements based on evidence
ASK THESE QUESTIONS:
1) What am I being asked to believe?
2) What evidence is available to support the claim?
3) Are there other ways of interpreting the evidence?
4) What other evidence would help to examine the alternatives?
5) What are the responsable conclusions?
Theory vs. Hypothesis
-THEORY explanation, can predict and suggest ways of controlling phenomena, can change or be added to with more research, general
EX) Eye movement activate parts of the brain in which bad experiences have been stored but not processed
-HYPOTHESIS clearly states what researchers think may be true, specific testable prediction
EX) EMDR treatment causes significant reduction to anxiety
Operational Defintion
- descriptions of the exact methods that will be used in the research
- ex) “EMDR Research” might be defined as 10 side to side eye movements per second
- to clarify hypothesis
Variable
- the specific things that are manipulated and measured in research
- ex) type of treatment
Data
- numbers or scores used to decide if a study’s results support the hypothesis
- have to consider the value of the evidence collected (is it reliable?)
4 Goals of Research
1) Describe behavior/mental processes
2) Make predictions about them
- naturalistic observation, case studies, surveys, correlations studies
3) Control them
4) Explain how/why they occur
- experiments
Naturalistic Observation
- process of watching as behavior occurs naturally
- can or cannot intervene
- but some problems -> for example, if ppl know they are being observed they act differently
IRB
- ethics institutional review board
- experiments must be approved by them
Informed Consent
-participants must know they are being researched and give their permission
Anonymity
-the participants identities must be kept secret
Right To Opt Out
-participant has the freedom to exit the study at any time
Debriefing
-participants must know the purpose of the study and can contact the researchers about study results
Social Desirability Bias
-tendency to answer questions based on what other ppl think is good/acceptable
Wording Effects
-the order of words/which words used can effect participants
Hawthorne Effect
-if you know you are being observed you act differently
Case Study
- intensive examinations of behavior/mental processes in a particular person or group
- useful when studying something new or rare
- can include tests, interviews
- unlikely to be representative of people in general
- can be used to test new treatments
Survey (Interview/Questionnaire)
- researchers ask people about their behaviors/attitudes/opinions
- provide a view of a large group of people, can get a lot of data
- how valid the survey is can depend on the wording of the questions, the types of people who answer the questions, and the social desirability bias
5 Steps in the Scientific Method
- scientific method = system of gathering data so error is reduced
1) recognize question & theory - what you think, explains
2) hypothesis - specific testable question
3) test hypothesis
4) draw conclusions
5) report results
Empirical
- provable by experiment
- quantitative data
Longitudinal Study
- a long term study of a sample
- thorough and accounts for change
- can cost a lot of money and sometimes participants opt out
Population
-total number of people
Random Sample
- selecting people from a population so that everyone has an equal chance of being chosen
- randomly selecting a sample from a larger population
Correlation Coefficient
- represents direction of relationship (+/-)
- represents strength (0 weak - 1 strong)
- 0.5 and above is considered strong
Scatter plots
- cluster of dots representing values of two variables
- slope = direction of relationship
- less scatter = stronger
Illusory Correlation/Third Variable
- correlation does NOT show cause and effect
- ILLUSORY CORRELATION: you think there is a correlation but there really is not
- THIRD VARIABLE: other factors can have an impact
Cohort Study
- a type of longitudinal study
- investigating a group with a certain trait for a long period of time
Cross Sectional Study
- Study lots of different people at the same time
- ex) all types of people with pHd’s
Measures of Variability
- RANGE: max-min
- VARIANCE: standard deviation ^2, degree of being spread out
- STANDARD DEVIATION: average distance from each data point to mean, square root of variance (smaller = more alike, larger = dissimilar)
- describe spread among scores
Skewed Distribution
- POSITIVE SKEW: mean > median > mode
- NEGATIVE SKEW: mode > median > mean
Normal Distribution
-mean median and mode are about equal
Histogram/bar graphs
-HISTOGRAM: looks like a bar graph with no gaps between classes, Y-axis is frequency
Frequency Polygon
- looks like a line graph
- used to compare or see the overall shape of the data
Limited deception only if necessary
- no deception unless justified by the study’s value
- deception that is essential to the experiment must be explained
Confidentiality
-psychologists must protect confidential information
Protection from all harm
-take steps to avoid harming patients and to minimize it if it is unavoidable
Correction
- if a study results in bad consequences it is the researchers responsibility to correct them
- unethical things must be fixed
Data set distributions/tables/frequency charts
-used to display data so conclusions can be made
Experimental vs. Control Group
- EXPERIMENTAL: the group that receives treatment
- CONTROL: group that receives no treatment, provide a baseline
- if everything about the groups was the same before the experiment, and changes should have been because of the treatment
Ind/Dep/Confounding Variables
- IND: the variable the researcher controls ex) which kind of treatment received
- DEP: variable affected by independent variable ex) anxiety level of patients
- CONFOUNDING: other favors that may confuse/effect the results ex) differences in treatment length
Random Assignment
- random process to decide if participants will be in experimental or control group
- distribute impact of uncontrollable variables across groups
- decreases the chance that these variables will change the results of the experiment
Placebo Effect
- improvements created by a participants expectations
- researchers normally have an experimental, control, and placebo group
- ex) sugar pills to help ppl quit smoking, they were helped a lot
Experimenter Bias
- unintentional effect experimenters may have on their results
- ex) experimenters may do a slightly better job administering treatment they think is better
Double/Single blind Study
- DOUBLE: when research participants and ppl giving the treatment don’t know who is receiving which type of treatment
- SINGLE: participants don’t know which group they are in
Sampling
-process of selecting participants for research
Representative Sample
-group of research participants who fairly reflect the characteristics of the population they were selected from
Measures of Central Tendency
- describe typical score
- mode, median, mean
ABD - Case Study and Longitudinal Study
- alike because both are in depth studies and are descriptive
- different because CASE STUDY studies a small group, while a LONG STUDY studies a group for many many years
ABD - Positive Skew and Negative Skew
- alike because median is in the middle
- different because POSITIVE SKEW has a longer tail towards the right and has the highest mean while NEGATIVE SKEW has a longer tail towards the left and has the highest mode
ABD - Correlation Study and Experiment
- alike because both discuss relationships between variables
- different because CORRELATIONAL has a goal of prediction while EXPERIMENTAL wants to explain/control
Statistically Significant
-outcome of a statistical test indicates that the probability of those results occurring by chance is small
Stratified Sample
- a Sample that is drawn from separate groups in a population
- ensures a more representative sample
Demand Characteristics
-participants know what is being investigated
Confirmation Bias
-when researchers try and find evidence that supports their hypothesis
Inferential Statistics
- help interpret meaning of descriptive statistics
- t-test: finds P (probability of error) want P to be low, measuring how likely findings are due to chance (degrees of freedom)
ABD - r and p
- alike because both help analyze/interpret data
- different because R = Correlational for strength and direction P = probability results are due to chance
ABD - random and stratified sample
- alike because both are randomly chosen and represent population
- different because RANDOM is chosen from whole population, and STRATIFIED is from specific groups