Research Methods Final Exam Study Guide

what is psychology?

"psyche" = mind "ology" = the study of *the study of the person*: -brain functions -perception, cognition, memory -behavior -mind-body connections -interpersonal relationships -psychopathology -the study of ourselves

how big should each block be?

# of conditions

social and cultural context of science

*the experimenter effects the experiment through the questions they ask, how they study them, how they interpret results, and how they are received.

basic format of APA style

-12pt standard font -1 inch margins -double spaced -typically no longer than 25-35 pages (15-20 for our class).

natural "treatments" or events

-9/11 -Katrina -No Child Left Behind

program evaluation

-Are human services effective? -Perhaps the most applied of all research -used to assess the effectiveness of human service organizations and provide feedback to administrators about their services -these evaluators assess needs, process, outcome, and efficiency of social services. -the relationship between basic research and applied research is reciprocal -true experiments and quasi-experiments can provide excellent approaches for evaluating social reforms -Key questions: --What are the needs? --What are the processes involved in program implementation? --What are the outputs? --What are the outcomes? --How efficient is the delivery of services?

references

-List everything you cited -Be careful of citing presentations, dissertations, and non peer-reviewed material ◦A study is only as good as the shoulders it stands upon

time series with nonequivalent control group

-Multiple pre and post-tests -Intervention and Control group -Rules out history and instrumentation effects

validity of quasi and true experiments

-Researcher as detective -Consider all threats to internal and external validity -Control and test for threats as much as possible -Note existing threats in limitations

single subject or (small-n) experiments

-Skinnerian analysis of behavior -apply experimental designs to a single subject or small group of single subjects (typically 5 or less) --manipulate an IV within the subject --examine changes in behavior ( DV) --it's possible to establish causal inferences in these experiments, because they're considered true experimental designs: -the DV=behavior change in the individual -the IV=the intervention that's been systematically manipulated *must begin with a stable baseline of DV to be able to detect a change

Tuskegee syphilis study

-The purpose of this study was to observe the natural progression of untreated syphilis in rural African-American men in Alabama under the guise of receiving free health care from the United States government. -The 40-year study was controversial for reasons related to ethical standards. Researchers knowingly failed to treat patients appropriately after the 1940s validation of penicillin was found as an effective cure for the disease that they were studying. -led to the requirement that institutions have an IRB

Null hypothesis significance testing (NHST)

-Use statistics to confirm if observed differences/relationships from a sample are likely to reflect true differences/relationships in the population -Uses probability theory to determine the likelihood that an effect would occur by chance --Limits ----Can't prove or disprove a hypothesis ----Can only provide probability statistics *this testing allows us to determine if the difference between groups is likely to be a true difference: -compare and evaluate group means and SD ...if the association between variables is likely to be a true association: -compare and evaluate correlations

confidence intervals

-a margin of error -95%-99% confident that the interval contains the true population mean = X bar plus or minus (t.05)(Sx) -where t.05 is drawn from the t-table by significant (95%-99%) and degrees of freedom (df= N-1)

confidence interval

-a margin of error -95%-99% confident that the interval contains the true population mean =mean plus or minus (t.05)(standard error of the mean) -Where t.05 is drawn from the t-table by significance (95%, 99%) and degrees of freedom (df = N - 1) *if 0 is not in between this than they're likely to be significant

requirements for a single subject experiment

-a potent intervention: must have an immediate effect on the DV -a specific behavioral DV -a stable baseline -controlled circumstances

what if I hypothesize that there is no difference between groups?

-a valid research question, but it's statistically problematic. -can never truly demonstrate that the null hypothesis is true--anything will be significant with large enough sample sizes --effect sizes

observation without intervention

-aka naturalistic observation. -the observer is a passive recorder. -observing behavior in natural settings with no manipulation- "the real world" -goal is to describe behavior as it naturally occurs.

characteristics of a survey

-all respondents complete the same items verbally or in writing -descriptive and predictive goals -correlational data about associations between variables: when you have correlation you can begin to make predictions. *used to obtain data about the feelings, attitudes, preferences, symptoms, etc. of a specific population of people. *you need a representative sample for data to be generalizable to a population.

limitations

-all studies have them -important to note them all -explain how the results can be generalized and how they can't -discuss how future research might correct current limitations.

alternatives to a no-treatment control

-alternate order of treatments -use of established treatments as the control -wait-list control -treatment as usual control

omnibus

-an effect in a complex design: do all variables together account for a significant portion of the variance in the DV?

participant observation

-an observer participates in a phenomenon to get an "inside view" -can be open or disguised: -disguised ex: policeman posing as a drug dealer selling standard drugs -can also ask participants in a study to observe their own behavior. -disguised participant observation is used to increase the likelihood of "natural" behavior. -"on being sane in insane places" Rosenhan 1973

random starting order with rotation

-another form of counterbalancing for incomplete designs -Start with a random order, then for each row, rotate one to the left to N rows -1st Order: C, D, B, A -2nd Order: D, B, A, C -3rd Order: B, A, C, D -4th Order: A, C, D, B

latin square

-another form of counterbalancing for incomplete designs -each condition appears in each ordinal position once -each condition precedes and follows each other condition exactly once

observational designs

-are done in a natural setting or a laboratory -are done with or without intervention -they must include: -precise definitions of conditions and behaviors -observations conducted in a systematic objective manner -careful and precise record keeping. -goal is to have full and accurate descriptions of behavior

randomly assigning to groups in independent groups designs

-balance out individual differences between groups -assumes the distribution of within-group differences are the same in each group -randomly assign the participants to levels of IV -not altering the assignment, must truly be random.

figures

-bar -line -best to demonstrate interactions

conclusions and implications

-brief (2-3 paragraphs at most) -broad importance of the study -application and implications for future research

multiple baseline design

-can be used in cases where reversal doesn't occur -compares the effect of an intervention on multiple baselines -multiple baselines can be across: 1. individuals 2. behaviors 3. situations **this won't work if an intervention generalizes across individuals, behaviors, situations

if distribution is not normal

-can cause problems with many statistical tests -may use a transformation or alternative statistics

external validity of multiple baseline design

-can sometimes generalize, sometimes not -can't test nonspecific, vague, or long-term interventions

disadvantages of the case study

-can't make causal inferences -observer bias -experimenter bias -threats to external validity- is it generalizable? Sometimes yes, sometimes no. -the danger of the testimonial

nonequivalent control group design

-compares an intervention to a "like" control group, but without randomization --the "likeness" is determined by a pretest -if groups are truly comparable, it controls for many threats to internal validity O1 x O2 ----------- O1 O2 -this design is vulnerable to: --Selection-history --Selection-instrumentation --Selection-regression to the mean --Selection-maturation --Regression --Expectancy, Contamination, Novelty effects

ABBA counterbalancing

-conditions in one sequence and then in reverse: -ABBA, ABCCBA, ABCDDCBA, etc. this isn't appropriate if: -practice effects are non-linear -there are anticipation effects -this is used for repeated measures complete designs

advantages of archival data

-data is already available -there are fewer ethical issues -less time consuming -less expensive

questionnaires

-demographic variables describe the characteristics of people who are surveyed -preferences and attitudes (either/or, yes/no, and Likert scale responses are the most common) -be careful of wording

descriptive statistics

-describe the distribution of the data -just as a sample is representative of the population, the sample statistics are estimates of true population values

inter-observer agreement & reliability

-do observers rate the same behavior in the same way? -if not: possible problems in coding scheme and observer training, which could make the data unusable

research results can differ by culture

-expectations for behavior can differ between and within cultures -historically, psychological research has sampled from narrow populations....but the results have been generalized to many populations -there's an increasing emphasis on representative sampling. --how far can the results generalize?

when to use the case study

-for new, rare, or unstudied phenomena or treatment -for initial data -to complement the nomothetic approach

stages of data analysis

-get to know the data: --correct errors --examine distributions -summarize the data: --descriptive statistics --graphs and tables -confirm results --statistical analyses -interpret results

true experiment

-has an intervention (one level of the IV) -has an appropriate comparison or control group (another level of the IV) -has a high degree of control (randomization, control of all conditions of the experiment) -use this if at all possible in order to make a clear cause-effect

threats to internal validity with an interrupted times-series design

-history, especially cyclical change --seasonal variations -instrumentation --new measures are often used with new programs/interventions -Maturation, testing, regression to mean are all well controlled by obtaining a baseline

controlling for bias

-instructions and training -blind experiments: really hard to do: not knowing what you should be paying attention to. -habituation: get used to it by being exposed to it: simply having the participants get used to you being around for a long time. Ex: neurons getting used to our clothes being on our bodies. -desensitization: actively doing something to help calm the situation

power and sample size

-just because an effect is not significant, doesn't mean it's not there -power= the probability that the null hypothesis will be rejected when it's false -power= the ability to detect statistically significant effects -Power = 1 - Type II Error (use power tables to determine based on α, df, n, and effect size) -Generally want power = .80 or larger

social experiments

-large scale, real-world tests -often more significant consequences -often greater investment of time, money, and effort -often driven by or affected by political concerns --ex: does adding extra police reduce crime? *outcome of social experiments cannot be "truly" evaluated

ethical concerns of participant observation

-loss of objectivity -observer could influence the behavior under observation -violation of privacy

continuous variables

-measures of central tendency -measures of dispersion -confidence intervals

observation with intervention

-most common type of observation in psychological research -used to: investigate the limits of a behavior, to study infrequent or normally inaccessible events, to study specific antecedent​ events, and to compare behavior under different conditions. -3 methods of this observation: 1. participant observer: observer participants in a phenomenon to get an inside view (can be open or disguised). Can also ask participants in a study to observe their own behavior. -structured observation: a data collecting method in which researchers gather data without direct involvement with the participants -field experiment: applies the scientific method to experimentally examine an intervention in the real world

interrupted times-series design

-multiple observations before intervention -multiple observations after intervention -comparison of baseline before and after intervention -simple interrupted times series design: O1 O2 O3 O4 O5 X O6 O7 O8 O9 O10 -time series with non-equivalent control group O1 O2 O3 O4 O5 X O6 O7 O8 O9 O10 -------------------------------------------- O1 O2 O3 O4 O5 O6 O7 O8 O9 O10 *Requirements -relatively stable baseline -a discontinuity in the time-series: must be abrupt

total number of blocks

-number required to balance out practice effects -calculate the average position of each condition over all trials where the average position for each is approximately equal

ethics of observational methods

-observing without knowledge of participants -privacy: is it really a public forum? -confidentiality -does your observation require informed consent? -how do you debrief? -how do you get the results to the participants?

difficulties conducting a true experiment in a natural setting

-obtaining permission from people in authority -obtaining participants --self-selected samples -randomization: --perception that randomization is unfair --ethics of the no-treatment control

advantages of incomplete designs

-often more feasible -An individual participant's responses cannot be interpreted: This is because the participant only receives one administration pattern such as condition A and then condition B, so order effects for this one participant rule out individual interpretation -At least one participant must complete each order in any design

statistical significance

-p < .05, .01, .001 -p < .10 a trend toward significance (use with great care) -Should report the exact p-value -The smaller the probability, the greater the chance that the effect will be replicated in another study -Effects should be replicated to be trusted

primary tool of survey research

-paper-and-pencil -structured and semi-structured interviews -internet *must be well-constructed, reliable, and valid.

ethics of archival data

-potentially less harmful alternative to direct procedures. -ethics of delving into records, public of private? -who gives permission? *always make sure that you're considering everyone's confidentiality

data collection

-qualitative: field notes and narrative reports -quantitative: coding sheets, physiological data -electronic recording

presenting descriptive data

-qualitative: presenting narrative reports with logical conclusions -quantitative: presenting nominal data: relative frequency (typically %'s) -interval/ratio data: typically use the mean (arithmitic average), median: middle point, mode: score that occurs most often

probability sampling

-random sampling -stratified random sampling (needs to be proportional to the population) **allows researchers to estimate the likelihood that their findings for the sample differ from those for the population. This is the method of choice for obtaining a representative sample.

social desirability

-response bias: are respondents telling the truth? -the respondents may be embarrassed, worried about consequences, or want to think of themselves in a certain way. -to improve truthful responses: -ensure confidentiality and anonymity -protect the data -follow instructions -follow a validity scale

ABAB design

-reversal design **must have reversal for this design: A= baseline B= intervention A= withdraw intervention (return to baseline) B= intervention **To make a causal inference: -behavior must change with 1st intervention -behavior must revert to baseline when the intervention is withdrawn -behavior must change again when the intervention is applied again

bias

-scientific writing should be free of implied or irrelevant evaluation of the group or groups being studied.

getting to know the data: step 2

-score the data: convert from raw to scored values -score questionnaires -create composite variables -should be planned out ahead of time

general APA style

-spelling and grammar -spell out numbers less than 10, use numbers for 10 and above. -use professional language (try for no I's and we's) -active vs. passive voice.

threats to internal validity

-testing intact groups (even when randomly assigning groups to different conditions) -extraneous variables -subject loss: mechanical or subjective subject loss -expectations (demand characteristics and experimenter effects)

alternative hypothesis (H1)

-the IV does have an effect on the DV -there is a difference between groups -there is an association between variables -ex: there is a difference between estimates and actual calories consumed

null hypothesis (H0)

-the IV has no effect on the DV -there is no difference between groups -there is no association between variables -ex: there is no difference between estimates and actual calories consumed

all possible orders counterbalancing

-the best choice for counterbalancing for incomplete designs -each participant is randomly assigned to one of all possible orders -problem is if there are N possible orders, there must be at least N participants

to be testable

-the constructs/variables must be adequately defined. -the hypothesis cannot be circular -a hypothesis must refer only to ideas recognizable to science.

reversal may not occur if

-the intervention generalizes -other variables cause the change in behavior (confounds) -other variables take control of the behavior after the intervention caused the change

interactions

-the main advantage of complex designs. Answers the question of whether the effect of one IV on the DV differs by the levels of another IV. -is the relationship between an IV and a DV the same for all groups? -are certain relationships true in some cases but not in all?

advantages of case studies

-they can be used to develop new ideas or hypotheses -theory development (they can act to challenge or support theories) -- "counterinstance" - a case that defies a generally accepted principle -they give us intensively detailed information, which is not usually available in the study of a group --idiographic approach (study of the individual) -can be used to try out new techniques or clinical innovations -can be used to examine very rare events (such as brain injuries, feral children, and Lesch-Nyan) -they can be more compelling and meaningful

advantages of multiple baseline designs

-they have high internal validity- very powerful for establishing causal inferences -they're useful in dismantling studies -- show what aspects of a treatment work -- show how they work

why are effect sizes important?

-they help determine if a statistically significant effect is meaningful in applied terms -useful in comparing effects across studies (meta-analysis) -necessary for power analyses/determining the necessary sample size for a study -can also be used in pilot data to determine if it's worthwhile to collect a larger sample

advantages of repeated measures designs

-they require fewer participants -they're greater sensitivity and power: it's easier to detect the effect of the IV -they can be more convenient and efficient -they can be better for studying behavior over time and when participants compare stimuli relative to one another

effect size and statistical significance

-they're NOT the same thing -Effect size is mostly independent of sample size, statistical significance is very much affected by sample size -Should report both, plus Confidence Intervals --Is the effect statistically significant? --How big is it? --How precise is the test?

threats to internal validity

-they're all controlled by a true experiment but not always by a quasi-experiment -the quasi-experiment is a compromise

disadvantages of a multiple baseline design

-they're limited to interventions with an immediate and specific behavioral effect -limited for behaviors with high variability --try to control for the variability --wait for stabilization --average the behavior

why should we use a quasi-experiment?

-to answer broader questions -when we want to examine broader populations -when we want to see if the intervention works in the real world/has external validity -b/c some hypotheses can't be tested with a true experiment

when not to use the case study

-to establish causal inferences -to try to make statements about groups -to try to prove a theory or test a hypothesis -to demonstrate the effectiveness of an intervention

step 4: design a method

-use operational definitions -choose measures: reliable, valid measures with previous support. If using a new measure, must have some way to support its validity. -choose a design: specific to your hypothesis. Multi-method designs are preferable. -specify your sample (human, animal, archival) and their specific group

tables and figures

-use sparingly -don't repeat info -refer to the reader the important results in a table/figure -be careful of misleading scale in tables and figures

matched groups

-used for small samples with large individual differences where randomization doesn't work well. -they can match participants across groups based on a relevant matching variable -they must have a good matching variable for example: -the DV -something related to the DV -or an important confound -randomly assign matched subjects to groups

tables

-useful for exact numbers -harder to interpret

effect size

-very VERY important! -tells us if any observed difference between groups is really substantive -an estimate of the size of the effect that is mostly independent of sample size -the p-value is very dependent on sample size -many types of these for different statistical errors: with a large enough sample size, any difference will be statistically significant so it's important to pay attention to these. -just because this isn't significant, doesn't mean it isn't there. -most common: Cohen's d and Pearson's r

disadvantages of archival data

-we must take what's there, we can't standardize how data is generated. -data may be subject to change over time

complex or factorial designs

1 DV with 2 or more IVs -variables can be nominal, interval, ordinal, or ratio -IVs can be independent groups and/or repeated measures: if one of each it's a "mixed design"

structure of a research report

1. abstract 2. introduction 3. method: participants, materials, procedure 4. results 5. discussion: limitations, implications, and conclusions. 6. references 7. notes (if applicable) 8. tables and figures (if applicable)

purpose of a research report

1. make a case 2. describe methods 3. summarize results 4. interpret and discuss what you did, why you did it, and what it means.

abstract

1. summarize: issue, method, findings, conclusions and implications 150-250 words -last thing you write, but ends up going first

determinants of power

1.Significance level= what you can actually see -smaller p-values mean less power (e.g. more power at .05 than at .01) -when you decrease the likelihood of type I error, you increase the likelihood of type II error 2. Effect size= what is there to be seen -it's easier to detect a large effect than a small one 3. Sample size= ability to see it/power -the most important determinant -The larger the sample size, the greater the likelihood of detecting a significant effect

Wilhelm Wundt

1879: "official" beginning of psychology as a science. He created the first psychology lab in Leipzig, Germany.

internal validity

3 conditions must be met for the experiment to have this and for the researchers to conclude that the IV caused the differences noted in the DV: -a time-order relationship -holding conditions constant -balancing

casual inference

3 important conditions: 1. covariation of events: if one event is the cause of another, the 2 events must vary together; when 1 changes, the other must change. 2. a time-order relationship: aka contingency: Ex: the headache was contingent on you hitting your head first. 3. plausible alternative events have been eliminated: you would have to rule out all other probable causes: ex: ruling out everything else that could have caused the headache.

Rosenhan's study of being "sane in insane places"

= an example of participant observation

N

= complete sample size

when lines are parallel

= no interaction effects

n

= sample size of each group and %

sample mean

= sum of all values/ sample size (N)

confidence interval to compare two group means

=(mean 1 - mean 2) plus or minus (t.05)(standard error of the mean 1 - standard error of the mean 2) -the standard error of the difference between means= √ [(n1 - 1)s1squared + (n2-1)s2squared][1/n1 + 1/n2] (n1 + n2 -if they're not intertwined it's likely they're significant

standard error of the mean

=the population standard deviation/ the square root of the sample size (N) =how good of an estimate of the population mean is the sample size

variance

=Σ(X - Xbar)2 / N-1 =standard deviation squared

step 3: develop a hypothesis

A testable statement about a phenomenon -Has at least two variables and specifies a relationship between them -May provide a causal explanation -Replication vs. New Data -Examples: -"Comfort eating" will be associated with greater weight -Hostile conflict will be associated with marital distress

which design is known for having high internal validity?

ABAB design

practice effects are non-linear

Abrupt initial changes followed by little change afterward -May ignore performance on early (and/or later) trials and attend to steady state -Easier to do with block randomization -Examples of non-linear effects: -Primacy and recency effects in memory -Initial changes in psychophysiological data

guideline 2

Be sensitive to labels: -describe participants as people, not labels -avoid portraying one group a "normal"

Alfred Binet

Binet-Simon scale of intelligence. He emphasized intelligence as a complex concept with multiple facets and subject to environment.

novelty effects

Can occur when an innovation, such as an experimental treatment, is introduced. -enthusiasm: ex: employees newfound enthusiasm, rather than the intervention itself, may account for the "success" of the intervention. -disruption effect: the opposite of a novelty effect: in which an innovation, disrupts employees' work to such an extent that they can't maintain their typical effectiveness. -Hawthorne effect (acting differently because you think people are watching and judging you)

guideline 1

Describe at the appropriate level of specificity. Use language that is accurate, specific, clear, and free of bias.

choosing the p-value

Do so before you begin a study -Avoid experimenter effects -Avoid inflating results -By probability, the more analyses you do, the greater likelihood of catching something significant...even if it's not a true effect -If using many analyses, set p at .01 or .001 instead of .05

non-probability (convenience sampling)

Ex: students in a class, first 50 shoppers when the mall opens -Easier to do, but results in biased sampling

example of natural groups

Hypothesis: Divorce will cause people to be more depressed than people who stay married -Problem with conducting an experimental study: Can't randomly assign people to a get divorced group or stay married group -Natural Group: Divorced group vs. Married group and assess levels of depression. -Could determine if two groups differed but couldn't say divorce caused depression

example of an independent groups design

Hypothesis: Drinking liquid will reduce appetite -Conditions (IV): ----Drink 1 glass of water prior to a meal ----No water prior to a meal -Participants randomly assigned to conditions, followed by a meal -Measure amount that's eaten (DV) -Individual differences will be evenly distributed across the groups (e.g. hunger, gender, weight, etc.) so group differences must be due to IV

example of a repeated measures design

Hypothesis: People will have a greater recall for emotion words matched to mood-congruent pictures than those matched to mood-incongruent pictures -IV: Mood congruence of words and pictures: -Emotion words matched to congruent pictures -Emotion words matched incongruent pictures -DV: Recognition of words after a 10-minute distracter task Conditions: -Congruent: Happy - Happy and Sad - Sad -Incongruent: Happy - Sad and Sad - Happy

example of matched groups

Hypothesis: Treatment A is better than placebo at decreasing depression -Problem: Researcher is conducting a pilot study with only 14 participants; there may be a lot of variability in severity of depression. -Solution: Assess participant's level of depression and randomly assign them based on their level of depression. The two most depressed are randomly placed in either Tx group or placebo group. Then the next two highest are randomly placed in either group and so on.

Milgram's obedience study

Milgram (1963) wanted to investigate whether Germans were particularly obedient to authority figures as this was a common explanation for the Nazi killings in World War II. -Milgram was interested in researching how far people would go in obeying an instruction if it involved harming another person. -He was interested in how easily ordinary people could be influenced into committing atrocities, for example, Germans in WWII. **used commission

test of probability for H0

NOT the probability that H1 is true -p-value aka a: --the probability of obtaining the observed effect if H0 were true --the probability of obtaining the observed effect by chance --Ex: Probability of finding the difference in our sample if, in the real world, there is no difference between estimated and actual calories consumed **Typically interested in p < .05...probability less than 5% that the observed effect would occur if there were no real effect in the population --There is still a chance that the observed effect is not a real-world effect

natural groups

Naturally occurring IVs/covariates, e.g., gender, ethnicity, abuse history, illness, etc. -Can't ethically or practically manipulate the IV -IVs are selected not manipulated -Compare natural groups -Not a true experiment -They may differ systematically -Can't make causal statements about these groups -Can manipulate other IVs and randomize within these groups

one group pre-test post-test

O1= pretest X= intervention or treatment O2= posttest -O1 x O2 *there's no control group at all *a bad choice for an experiment

mechanical subject loss

Occurs when a subject fails to complete the experiment because of equipment failure or because of experimenter error. -wasn't planned, it just happened.

selective subject loss

Occurs when subjects are lost differentially across the conditions of the experiment as the result of some characteristic of each subject that is related to the outcome of the study. -the person selected not to be a part of the study; not random; more serious because it may directly affect the study

example of random starting order with rotation

Present four television programs to participants and have them rate their mood after watching each one -Randomly assign participants to watch programs in one of four orders: 1. Sad, Exciting, Violent, Happy 2. Exciting, Violent, Happy, Sad 3. Violent, Happy, Sad, Exciting 4. Happy, Sad, Exciting, Violent *used for repeated measures incomplete designs

identifying main effects

Same as in a simple design - the overall values for the DV at each level of the IV -Confirm with statistics

additive effects with selection

Selection-history Selection-maturation Selection-instrumentation Selection-regression to the mean

the research process

Step 1 - ask a question Step 2 - study the literature Step 3 - develop hypotheses Step 4 - design a method Step 5 - collect data Step 6 - enter and analyze data Step 7 - interpret data Step 8 - disseminate results

Stanford prison study

The Stanford prison experiment was an attempt to investigate the psychological effects of perceived power, focusing on the struggle between prisoners and prison officers. It was conducted at Stanford University between August 14-20, 1971, by a research group led by psychology professor Philip Zimbardo using college students.[1] It was funded by the U.S. Office of Naval Research[2] as an investigation into the causes of difficulties between guards and prisoners in the United States Navy and United States Marine Corps.

why should we care about interactions?

They provide information about the limits of the effects of an IV. -if there is no interaction, the results may be generalizable to a larger group -if there is an interaction, the results are limited to a specific group and are less meaningful They also provide support for theories. They help us understand group differences. *real effects can be hidden if an interaction is not assessed.

Hashemi & Cochrane (1999)

They studied expressed emotions (EE) and schizophrenia outcomes. -EE is the critical, hostile, over-involved attitude that relatives have towards the person with a disability. -20 British Sikh, 20 British Pakistani, 20 British White -High EE: 80% of Pakistanis, 45% of Whites, 30% Sikhs -EE predicted relapse for Whites but not for Pakistanis or Sikhs

step 7: interpret the data

Was the hypothesis supported? To what degree? If not, why not? What do the results mean? What do they imply for future research? How can they be applied? -Probability vs. Exact Answers -Danger of going beyond the data

spurious relationship

a correlation that can be explained by a third variable

theory

a logically organized set of propositions (claims, statements, assertions) that serves to define events (concepts), describe relationships among these events, and explain the occurrence of these events. Ex: for a " " of flashbulb memory, it must state exactly what a flashbulb memory is and how it differs from typical memories.

scientific method

a means to gain knowledge. It refers to the ways in which questions are asked and the logic/methods used to gain answers. -two important characteristics are an empirical approach and a skeptical attitude.

block randomization

a method of randomization that assigns subjects within a block -each block contains all conditions in random order -2 x the number of groups -size of block= number of conditions -number of blocks= number of times each condition is administered -uses groups of equal size -this type of randomization is useful for time-related and group-related variables in randomization -this is used for repeated measures complete designs -use this randomization when you have smaller numbers and need equal groups.

informed consent

a person's explicitly expressed willingness to participate in a research project based on a clear understanding of the nature of the research, and of all factors that might be expected to influence that person's willingness to participate. -Written consent is absolutely essential when participants are exposed to more than minimal risk. -In some situations, researchers aren't required to obtain informed consent. Ex: when observing individuals behavior in a public place without any interference.

variable ratio schedule

a reinforcement schedule that reinforces a response after an unpredictable number of responses

variable interval schedule

a reinforcement schedule that reinforces a response at unpredictable time intervals

fixed interval schedule

a reinforcement schedule that reinforces a response only after a specified time has elapsed

covariation

a relationship between the IV and the DV in an experiment.

time-order relationship

a relationship that is measured by manipulating the IV and then seeing the difference in DV

step 1: ask a question

a research question should be: -relevant -important -have practical implications

selection bias

a specific group within a population is under or over-represented. Ex: whites tend to be over-represented while minorities tend to be under-represented. Middle-high SES tends to be over-represented while low SES tends to be under-represented.

minimal risk

a study is described as involving "minimal risk" when the procedures or activities in the study are similar to those experienced by participants in their everyday life.

Tearoom trade study

a study of homosexual encounters in public places. The study is an analysis of male-male sexual behavior in public toilets. Because the researcher misrepresented his identity and intent and because the privacy of the subjects was infringed during the study, this study has caused a major debate on privacy for research participants and is now often used as an example of highly controversial social research. **used commission

monster study

a stuttering experiment performed on 22 orphan children in Iowa. Half of the children received positive speech therapy, praising the fluency of their speech, and the other half, negative speech therapy, belittling the children for speech imperfections. Many of the normal speaking orphan children who received negative therapy in the experiment suffered negative psychological effects and some retained speech problems for the rest of their lives. -one of the most unethical studies ever done

the risk/benefit ratio

a subjective evaluation of the risks and benefits of a research project is used to determine whether the research should be conducted. Asks the question "Is it worth it?" and "are the benefits greater than the risks?". If the risks outweigh the benefits, the IRB will not approve the research but if the benefits outweigh the risks, they will approve it.

data analysis

a subsection of the Method section that should only be included if your statistical approach is particularly complex and not a common way of examining the info you gathered

selection-instrumentation effect

a threat due to the combination of selection and instrumentation occurs when changes in a measuring instrument are more likely to be detected in one group than they are in another.

testing

a threat to internal validity because taking a test could have an effect on all subsequent testing

regression

a threat to internal validity. Extreme scores regress toward their mean

instrumentation

a threat to internal validity. Refers to changes in your instruments: could have an impact

subject attrition

a threat to internal validity. Refers to selective and mechanical subject loss

maturation

a threat to internal validity. The change associated with the passage of time-which could have an impact

history

a threat to internal validity. The history or past experiences of participants can have an impact

selection

a threat to internal validity. The placement of participants in groups not random

use traces

a type of physical trace that reflects the physical evidence of use (or nonuse) of items and can be measured in terms of natural or controlled use. Ex's: remains of cigarettes in ashtrays, aluminum cans in a recycling bin, and litter on a campus walkway. Clock settings are a _____ that may tell us about the degree to which people in different cultures are concerned with punctuality, and marks in textbooks may inform researchers which topics students study the most.

construct validity

actually measuring what the theory says you're measuring -convergent: measurements that should be related are related (i.e. they correlate highly) -divergent (discriminant): measurements that should be unrelated actually are unrelated (i.e. they do not correlate highly with one another)

independent groups designs

aka between-subjects designs. This design allows researchers to compare differences between groups (the effect of the IV) while controlling for differences within groups (individual differences or error) -There's a separate group of participants for each level of an IV. Ex: randomly assigning 20 subjects to the IV group and 20 subjects to the control group. -Can have multiple groups, usually limited to 3-5 at most

introduction

aka literature review-why you did this study. -brief- comprehensive but not exhaustive -makes a case for the hypotheses and the study- not just a summary of previous research. -describe the issue: why is it important? -theoretical implications, summary of background literature with citations -purpose, rational and general design of study -present hypotheses -must be justified by the literature.

measure of dispersion

aka variability: indicate the breadth/variability of the distribution -the crudest measure of dispersion is the range: represented by the lowest and highest scores in the distribution.

population

all people within a group being studied

quasi-expirment

an "almost" experiment/a compromise -field experiments and independent groups-natural groups are examples of this kind of experiment -differentiated from a true experiment because there is a lack of control in this experiment -has an IV and DV -includes a comparison group -lacks randomization and tight control of experimental conditions

selection-maturation effect

an additive effect of selection and maturation occurs when individuals in one group change at a different rate than individuals in another group. Ex: individuals in one group may grow more experienced, more tired, more bored, or less healthy at a faster rate than individuals in another group.

EAR

an electronically activated recording device which provides an acoustic log of a person's daily activities.

quasi-experimental

an experimental design that lacks random assignment

application

answers the questions: -how can this information be used to affect the phenomenon? -how can basic research be applied to improve lives?

prediction

answers the questions: -what is associated with the phenomenon? -what can it lead to? -what leads to it?

explanation

answers the questions: -why does the phenomenon occur? -what causes it?

description

answers the questions: -What is the phenomenon? -What are the primary characteristics? -How do we know when it is present? -For whom does it occur? -Under what circumstances does it occur?

in an independent groups, random design block randomization is often used to randomly assign participants to groups. What is the block size?

any multiple of the number of conditions (i.e. groups) but usually 2x the number of conditions

representative sample

are "like" the larger population from which they're drawn.

products

are the creations, constructions, or other artifacts of behavior. Ex: by examining the types of vessels, paintings, tools, and other artifacts, anthropologists can describe patterns of behavior from thousands of years ago. Modern day ____such as TV, music, fashion, and electronic devices can provide insight into our culture and behavior.

external validity

as usual, this is related to the representativeness of the sample --do the effects generalize to the larger population? --are the effects of the intervention the same or different for different groups? *replication is the best test for external validity

random assignment

assigning participants to experimental and control groups by chance

error

because NHST relies on probability, there is always the chance for error -hypotheses are supported or not supported but never proved or disproved

power analyses

before you begin a study, you should conduct this to determine the necessary sample size -need to know the effect size of interest -need to know the planned p-value -Fairly easy with simple analyses, but can get very complicated with more sophisticated analyses

selective deposit

bias in the original production of the data: what did you actually put in? Ex: on facebook we select which photos to post. We only post ones we like. -* what was put into the archive

deception

can occur either through omission, the withholding of info, or commission, intentionally misinforming participants about an aspect of the research. -researchers must carefully weigh the costs of deception against the potential benefits of the research when considering the use of deception.

contamination

can occur in a true experiment when there is communication of information about the experiment between groups of participants. Several unwanted effects can occur when groups communicate information about an experiment: -resentment: is possible when individuals in a control condition learn they're not receiving a desirable treatment -rivalry: rather than responding with poorer performance, individuals in a control condition may feel motivated to work harder so they don't look bad compared to individuals in the treatment group. -diffusion of treatments: if treatment info is communicated between groups (e.g. during lunch breaks), individuals in a control condition may apply the info to themselves to imitate those receiving the treatment *experimenter expectancy: when an experimenter unintentionally influences the results

differential regression

can occur when regression is more likely in one group than in another.

ethnocentric bias

can occur when researchers fail to recognize when experiences and values of their own life/culture affect their interpretations of behavior observed in other cultures.

step 6: enter and analyze data

choose appropriate analytic technique

getting to know the data: step 1

clean the data: look for and correct errors -double entry -data checking -examine ranges for each variable outliers: -very high or very low data points that appear unusual -they may be a result of error, they may be a real value -can delete if they're an error, but note and justify

William James

considered to be the "father" of American psychology. Developed the technique of introspection to investigate mental processes.

internal consistency reliability

consistency across items within a test: are they all measuring the same thing? (usually measured using a statistic called Chronbach's Alpha for continuous or dichotomous​ variables.

data coding

conversion of observed behavior into quantitative data. -scales: 1. nominal: sort into categories. Ex: red cars and blue cars 2. ordinal: rank-order Ex: places in a race: 1st, 2nd, 3rd, etc. 3. interval: equal distance between points on a scale: Ex: IQ tests, SAT/ACT scores 4: ratio: interval scale with an absolute zero: Ex: clocking time, money

criterion-related validity

correlation between the test and some criterion, such as another test or an outcome -Concurrent - measured at the same time -Predictive - measured at some time in the future: ex: using law school GPA to determine a correlation with LSAT scores.

balancing: incomplete design

each participant receives or is administered each condition only once. Types of counterbalancing: -Use all possible order conditions -Latin square -Random starting order with rotation -Practice effects balanced across rather than within participants -Conditions applied to participants in different orders -Each condition must appear in each ordinal position equally often

balancing: complete design

each participant receives or is administered each condition several times in different order. -The conditions are administered enough times to balance out practice effects. -practice effects are balanced within individuals Types of counterbalancing: -Block randomization -ABBA counterbalancing

example of an interaction effect

easy going parents were more similar to the assertive parent in reporting problems with the anxious child, but they were much more similar to the aggressive parent when reporting problems with the ADHD child.

empirical approach

emphasizes direct observation and experimentation as a way of answering questions.

simple random sample

every element has an equal chance of being included in the sample.

Likert-scale

ex: on a scale of 1-7 rate how satisfied you are with the class.

culture

expectations for behavior differ between and within cultures.

operational definition

explains a concept solely in terms of the observable procedures used to produce and measure it.

Sir Francis Galton

first attempts to measure intelligence. He demonstrated that it is hereditary, but needs environmental support.

Sigmund Freud

focused on understanding personality, mental disorders, and the unconscious using his method of free association.

ethics code

formulated by the APA to set standards of ethical behavior for psychologists. Ethical decisions are best made after consultation with others, especially those who have expertise in the field of study.

multimodal distribution

graph with multiple peaks

bimodal distribution

graph with two peaks

sample

group sampled from the population

reliability

how consistent a measure is. How consistent is the instrument being used and the observations being made? If two or more observers agree then the observation is considered " ".

example of a simple main effect

in the depressed children condition, parents with an assertive style reported the fewest behavior problems and parents with an aggressive style reported the most behavior problems

field notes

include only the observer's running descriptions of the participants, events, settings, and behaviors that are of particular interest to the observer, and may not contain exact records of everything that occurred. They're used by journalists, social workers, anthropologists, psychologists, and others and are probably used more frequently than any other kind of narrative record. -they tend to be highly personalized. Ex: a clinical psychologist may record specific behaviors of an individual with knowledge of that individual's specific diagnosis or particular clinical issues.

descriptive methods

include: observational designs, surveys, and archival/trace data. These methods do not include controlled manipulation of a variable (experimentation).

demand characteristics

individuals often react to the presence of an observer by trying to behave in ways that they think the researcher wants them to behave.

Ivan Pavlov

inventor of classical conditioning. -Unconditioned Stimulus (meat powder) -Unconditioned Response (dog salivates) -Pair US and Conditioned Stimulus (bell) -Conditioned Response (dog salivates)

B.F. Skinner

inventor of operant conditioning. -Operant (rat presses a bar) -Reinforcing stimulus (food pellet drops)

situation-sampling

involves studying behavior in different locations and under different circumstances and conditions. This kind of sampling enhances the external validity of findings. By sampling various situations, researchers reduce the risk that their results will be unique to specific conditions. Ex: animals don't behave the same in zoos as they do in the wild

construct

is a concept or idea; examples of psychological " " include intelligence, depression, aggression, and memory.

hypothesis

is a tentative explanation for a phenomenon. Ideally, it's designed to generate knew knowledge or replicate previous findings. -based on scientific evidence, not feelings, personal experience, or intuition

debriefing

is necessary to explain to participants the need for deception, to address any misconceptions participants may have about their participation, and to remove any harmful effects resulting from the deception. -the goal of this is to educate the participants about the research and to leave them feeling positive about their participation in the study.

stem and leaf display

is particularly useful for visualizing the general features of a data set and for detecting outliers

a discontinuity in the time series

is the major evidence of an effect of treatment

coefficient of determination (r2)

is typically used to determine the percentage of variation in the DV due to the IV -can occasionally be used as an effect size

scientific reporting

is unbiased and objective; uses operational definitions.

control

it is the essential ingredient of science, distinguishing it from non-scientific procedures. By using controlled observation, scientists gain a clearer picture of the factors that produce a phenomenon.

sampling frame

list of people in the population

getting to know the data: part 3

look at it: -stem-and-leaf displays and histograms -notice outliers -begin to observe group differences normality of the distribution: -very important because many statistics assume a normal distribution

content analysis

making inferences based on objective recordings of archival data -identify the relevant source: is it good? -use operational definitions -sample selections: is it representative of the population? -code the data: is it reliable and valid?

structured observation

manipulations to elicit a behavior -disguised through the use of confederates: individuals involved in the study who are instructed to behave in a certain way (Ex: milgram's experiment) -overt instructions and then observe responses, such as Piaget's studies. *grocery stores are all modified using this kind of observation to try and get you to buy more: milk is always in the way back and the candy is always at the register. -* modifying one thing: ex: skipping across campus instead of walking. Only changing one condition to see its effect. -inattentional blindness study

subject-sampling

may be used to examine only some individuals in a setting. Ex: during peak lunchtime, it would be impossible to study all the students in the cafeteria, so researchers would use this type of sampling to systematically select students. The goal is to obtain a represenattive​ sample.

measures of central tendency

mean: mathematical average median: the middle value mode: the most common/frequent value

Hermann von Helmholtz

measured the speed of the neural impulse: 90ft per second. (1821-1894)

test-retest reliability

measures consistency across time (measured by correlating the test with the retest using the Pearson r)

parallel forms reliability

measures consistency within forms or across similar forms (tests, questionnaires, etc)

validity

measuring what you think you're measuring (corresponds accurately to the real world) -3 primary types of measuring the validity of a test or questionnaire: 1. construct 2. content 3. criterion-related

selective survival

missing or incomplete data. The data that survived over the years. There's a chance that data was lost.

sample size

more levels/more IVs means more complexity...but results in smaller cell sizes and reduced power -less of a problem in repeated measures designs

an archeologist primarily uses

natural use physical trace data

content validity

non-statistical validity, usually done by experts who agree the content fits with the scientific literature

bias

observer influences the observed: -reactivity: physiological reactions, social desireability observed influences the observer: -change in perceptions expectancies

reactivity

occurs when people react to the fact that they're being observed by changing their normal behavior. This is why researchers may choose to disguise their role as observers

interviewer bias

occurs when the interviewer only records selected portions of the respondents' answers or tried to adjust the wording of a question to "fit" the respondent.

anticipation effects

occurs when the participant perceives a pattern and changes their response accordingly -particularly likely with an abba design

confounding

occurs when two potentially effective independent variables are allowed to covary simultaneously. When research is " ", it's impossible to determine what variable is responsible for any obtained difference in performance.

type 2 error

occurs when you fail to reject the null hypothesis when it's actually false -you fail to find statistical significance for a real effect

type 1 error

occurs when you reject the null hypothesis but it's actually true *when you think something is going on, but in the real world, nothing is happening -you find statistically significant results to support your hypothesis, but the effect isn't real -this is a problem because we often want to believe we're right so we're inclined to look for supporting evidence even when it's not there. It's hard to accept being wrong so this error occurs because we're so set on being right.

cross-sectional survey design

one or more samples from a population at one time -slice the population into sections and study all the sections at the same time. Ex: taking a bunch of 1st, 2nd, 3rd, 4th, etc. graders and comparing their attitudes about school. -one or more samples are drawn from the population(s) at one point in time. - these designs allow researchers to describe the characteristics of a population or the differences between two or more populations -correlational findings from these designs allow researchers to make predictions.

two possible outcomes

p < .05: Reject the Null Hypothesis -Supports the alternative hypothesis (the study hypothesis) p > .05: Fail to reject the Null Hypothesis -Does not support the alternative hypothesis -Cannot prove either H0 or H1

behavior sampling

parallel to sampling of people or animals -time sampling: observe at specific time intervals -event sampling: observe at a specific event -situation sampling: observe in many different locations under different circumstances -subject sampling: sample select subjects in a situation

differential transfer

performance on one condition is dependent on the condition that precedes it (i.e. the first condition influences all subsequent conditions) -instructions -interventions -can't use repeated measures design -Test for this by comparing results of repeated measures and independent groups

response rate bias

poses a threat to representativeness when participants fail to complete a survey

determining risk

potential risks in psychological research include: risk of physical injury, risk of social injury, and risk of emotional/mental stress. -risks must be evaluated in terms of potential participants' everyday activities, their physical and metal health, and capabilities.

step 8: disseminate the results

present and/or publish the data (peer review, quality of outlet). -response of the scientific community -A single study is insufficient to answer any research question - hypotheses and theories are supported by multiple studies, conducted by multiple people, in multiple ways

qualitative research

produces verbal summaries of research findings with little statistical summaries or analysis.

correlational research

provides a basis for making predictions. Relationships among naturally occurring variables are assessed with the goal of identifying predictive relationships.

applied research

psychologists conduct research in order to change people's lives for the better.

archival data

public and private records for individuals, organizations, countries, etc. -running records -media -other records *if you generate the data, it's not this. You must be using past data. This data seldom answers questions for us, but rather it gives us more ideas to support our own research.

quantitative research

refers to studies in which the findings are described using statistical summary and analysis.

validity

refers to the "truthfulness" of a measure. A " " measure of a construct is one that measures what it claims to measure.

reliability

refers to the consistency of the data: are similar results produced under consistent conditions? -test-retest: measures consistency across time (measured by correlating the test with the retest using the pearson r) -parallel forms: measures consistency within forms or across similar forms (tests, questionnaires, etc) -internal consistency: consistency across items within a test: are they all measuring the same thing? (usually measured using a statistic called Chronbach's Alpha for continuous or dichotomous​ variables.

interobserver reliability

refers to the degree to which two (or more) independent observers agree. When observers disagree, we become uncertain about what is being measured and what behaviors/events actually occurred.

nomothetic approach

refers to the discovery of general scientific laws. Psychologists try to establish broad generalizations of behavior that apply to a diverse population.

external validity

refers to the extent to which the results of a research study can be generalized to different populations, settings, and conditions.

publication credit

refers to the process of identifying as authors those individuals who have made significant contributions to the research project.

privacy

refers to the rights of individuals to decide how info about them is to be communicated to others. -researchers should consider the sensitivity of the info, the setting, and the method of dissemination of the info.

fixed ratio schedule

reinforcement is delivered after a specific number of responses have been made

physical traces

remnants, fragments, and products of past behavior. Ex: blood splatter, bones, etc. -use traces: -natural use: no intervention, naturally occuring. -controlled use: some degree of manipulation

best evidence for external validity

replication with different populations, settings, and times

narrative records

researchers often use these when they seek a comprehensive record of behavior. They provide a more or less faithful reproduction of behavior as it originally occurred. To create this, an observer can write descriptions of behavior, or use audio or video recordings. Ex: videos were used to record the mother-child interactions among maltreating and non-maltreating families. -should be made ASAP after behavior is observed

basic research

researchers seek primarily to understand behavior and mental processes for its own sake.

Institutional Review Boards (IRBs)

review psychological research to protect the rights and welfare of human participants.

Institutional Animal Care and Use Committees (IACUCs)

review research conducted with animals to ensure that animals are treated humanly.

van Ijzendoorn & Kroonenberg (1988)

reviewed 32 studies of child attachment across 8 countries. -They found strong cross cultural differences. -Secure Attachment predominant across all countries -Western European samples had greater Avoidant Attachment -Japan and Israel tended had greater Anxious Attachment -even stronger within culture differences.

repeated measures t-test

same subjects being tested under different conditions

experiment

scientists manipulate one or more factors and observe the effects on behavior.

multi-method approach

searching for an answer using various research methodologies and measures of behavior. *Using multiple methods "fills in all the gaps" if any individual method is flawed.

literature review

section of the research report where you pull material together to generate knew hypotheses and make a case for why they should be tested.

nomothetic approach

seeks to determine the typical or average of a group

pre-selection bias

selecting only certain data: what you've decided to use from the archives. Ex: deciding to use newspapers from specific cities.

successive independent samples

series of cross-sectional designs on the same population, but a different of respondents, over time. In this design, different samples of respondents from the population complete the survey over a time period. -samples must be drawn from the same population and the same questions must be asked. -this design allows researchers to study changes in a population over time. - it does not allow researchers to infer how individual respondents have changed over time. -A problem with this design occurs when the samples drawn from the population are not comparable-that is, not equally representative of the population.

element

single person from the sample

computer programs

statistical programs are highly beneficial (and necessary with more complicated analyses) -SPSS & SAS among the most popular in psychology BUT ...you still need to know the stats and the formulas behind them

behaviorism

stresses that psychology should attend to only what is observable: behavior.

idiographic approach

study the individual rather than groups. These researchers believe that although individuals behave in ways that conform to general laws or principles, the uniqueness of individuals must also be described.

non-probability sampling

such as convenience sampling, does not guarantee that every element in a population has an equal chance of being included in the sample.

longitudinal survey design

taking the same group and studying them over time. -takes way longer -the same respondents are surveyed over time in order to examine changes in individual respondents. -Because of the correlational nature of survey data, it is difficult to identify the causes of individuals' changes over time. -As people drop out of the study over time (respondent mortality), the final sample may no longer be comparable to the original sample or represent the population.

analysis of variance (ANOVA)

test for differences between 3 or more groups or levels of the IV (can be used in independent groups, repeated measures, or complex designs)

t-test

tests for differences between 2 independent groups -repeated measures t-test

randomization in independent groups designs

the assumption with this is that random assignment will result in equal distributions of individual differences across groups. -valid with large, homogeneous groups -affected by the distribution of individual differences within the population -need to test the validity of the assumption -this can fail, especially when the sample size is small which can result in unequal group sizes.

Pearson's r

the correlation between two variables (correlation coefficient) ΣXY - [(ΣX)(ΣY)]/N divided by √[ΣX2 - (ΣX)2/N][ΣY2 - (ΣY)2/N] -x= raw score on first variable -y= raw score on second variable -ΣXY= sum of cross products of X and Y (multiply X and Y and sum the products) -N= number of pairs of scores -can use this to determine the effect size Small:r > .10 Medium:r > .30 Large:r > .50 -Some use the Coefficient of Determination (r2) to obtain an effect size, although more typically it is used to determine the percentage of variation in the DV due to the IV

repeated measures design

the design where each participant completes/is exposed to all the conditions in an experiment. -the participants serve as own controls which eliminates within groups individual differences

continuous reinforcement schedule

the desired response is reinforced every time it occurs

applied research

the examination of psychological principles and treatments in real-world settings. Includes: -case studies -single-subject experiments -quasi-experimental designs -clinical trials -program evaluation

plagiarism

the fact that both professionals and students commit acts of plagiarism suggests that many people too often veer from the tightrope by seeking their own recognition instead of giving due credit to the work of others. **Cite even when paraphrasing.

time-sampling

the goal of this sampling is to obtain a sample of behavior that will represent an organism's usual behavior.

the case study

the intensive description and analysis of a specific case -it doesn't use experimental techniques -there's a low degree of control -it's based on clinical impressions -it can include empirical data (info acquired from observation or experimentation) *it's an exploratory method *it contributes to the development of hypotheses and theories -these hypotheses and theories can't be tested directly -they can sometimes be disproved

practice effects

the major disadvantage of repeated measures designs. They're the changes that are produced because of repeated testing or exposure (not because of IV as we want) -Threaten internal validity - We can control these through balancing or counterbalancing

mean

the mathematical average

median

the midpoint/middle value. If you have a skewed distribution this may be better to use than the mean.

the larger the sample

the more accurate the estimate

mode

the most common value. If you have multiple of these, you can't have a normal distribution.

standard deviation

the most commonly used measure of dispersion (the counterpart of the mean) -tells you approximately how far on the average a score is from the mean =square root of the variance

conditions

the number of these = the product of the number of levels in each IV. Ex: -2x2= 4 ________ -3x3= 9 ________ -3x4x2= 24______

stratified random sample

the population is divided into subpopulations called strata and random samples are drawn from each of these strata. The representativeness of a sample can often be improved by using this kind of sampling. Also useful for when you want to describe specific proportions of a population.

power

the probability that the null hypothesis will be rejected when it's false -the ability to detect statistically significant results

data reduction

the process of abstracting and summarizing behavioral data. Ex: in qualitative data analysis, this occurs when researchers verbally summarize info, identify themes, categorize and group pieces of info, and incorporate their own observations about the narrative reports.

Cohen's d

the size of the difference between the means of two groups d = (xbar (aka mean)1 - Xbar2)/SDpop SDpop = √ (n1 - 1)s1squared + (n2 - 1)s2squared/ N -n1=sample size of group 1 -n2= sample size of group 2 -s1squared=variance of group 1 -s2squared=variance of group 2 N=n1+n2 -this is determined by: --the size of the difference between the two groups --the amount of variability within the groups Small:d > .20 Medium:d > .50 Large:d > .80

f-test

the statistic used in an ANOVA

idiographic approach

the study of the individual

observer bias

the systematic errors in observation that result from an observer's expectations. Ex: in Rosenbaum's study, once the observers were labeled schizophrenics, staff members interpreted their behavior solely according to this label.

multi-method approach

the use of several research techniques in the same research project -recommended because it reduces the likelihood that research findings are due to some artifact of a single measurement process -vital to research

in which of the following examples would it be most reasonable to select a repeated measures, incomplete design?

there are 3 videos to show participants

experimental designs

these methods allow us to determine cause and effect.

types of complex designs

they're described by the number of IVs and the number of levels/conditions in each IV -simplest: 2x2: 4 conditions: 1 interaction: 2 main effects -more complex: 2x2x2, 3x3, 3x4x2, etc.

unobtrusive measures

they're nonreactive. They can be obtained by examining physical trace and archival data.

field experiments

they're the most extreme form of intervention in observational studies. -researchers exert more control -researchers often manipulate an IV to see its effect on behavior. Ex: confederates posing as robbers study. -* requires that you've set up 2 conditions and are comparing the conditions effects.

selection-history effect

this problem arises when an event other than the treatment affects one group and not the other.

purpose of APA style

to present research in a clear and unbiased manner.

natural use traces

traces that are observed without any intervention by a researcher and reflect naturally occurring events

controlled use traces

traces that result from some intervention by a researcher. Ex: colored vs. non-colored potato chips. Researchers intervened or altered the stack of colored potato chips.

step 5: collect data

train staff to administer the tasks and get people to participate fully and honestly.

the most important factor in reducing observer bias is

training the observers

what is the scientific method?

understanding the world through empirical evidence rather than intuition. -Goals: description, prediction, explanation, application.

event-sampling

used to describe events that happen infrequently. Ex: observing an animal's behavior while they're eating. The particular event defines when observations are to be made. Also useful for observing behavior during events that occur unpredictably, such as natural disasters. -can lead to convenience sampling: observers may sample at times that are most "convenient"

meta-anlysis

using data from a bunch of different studies to evaluate something across them all.

main effects

what are the effects of the individual IVs?

results

what happened in the study. Present and summarize analyses-don't interpret yet. -include: -purpose of the analysis -type of analysis -summary of analytic results, including reference to any tables and figures. -present confidence intervals, effect sizes, etc. -summary statement

step 2: study the literature

what is known about this topic already? Reviews existing research and theories critically.

simple main effect

what is the effect of one IV at one level of a second IV?

discussion

what it means. -Summarize results briefly ◦Which hypotheses were or were not supported?◦Were the results as expected or were there surprises? -Discuss the meaning/interpretation of results ◦What is the meaning of these specific results? ◦How do the results generalize beyond the sample (if they do)? ◦What do the results mean for science and theory more broadly? -Be careful not to go beyond the data◦Causal statements ◦Excessive theorizing

method

what you did and how you did the study. -participants: -how many?, where did the data come from (human, animal archival)?, what were the characteristics of the participants (age, species, source, etc.)? -materials: -what kind of data was collected? -procedure -how was the data collected?

kurtosis: leptokurtic

when the graph gets higher/narrower around the mean

skew negative

when the graph gets lower and lower towards negative values

skew positive

when the graph gets lower and lower towards positive values

kurtosis: platykurtic

when the graph gets wider/starts to plateau

ceiling effects

when the performance on a DV reaches a maximum -can't interpret interactions

floor effects

when the performance on a DV reaches a minimum -can't interpret interactions

correlation

when two variables are associated with one another.

can you have too much power?

yes! -with a large enough sample size, any difference will be statistically significant --increases the likelihood of type I error -another reason to pay attention to effect sizes -we generally worry more about type I error than type II -p < .05 -Power only needs to be .80 -As a result, Type II error is much more common

Rosenfarb, Bellack, & Aziz (2006)

•Studied Expressed Emotion (EE) among African American and Caucasian patients with schizophrenia and their families •Interaction effect for Ethnicity and EE: -Caucasian patients did worse when family members were critical/intrusive and the patient had odd thoughts -African American patients did better when family members were critical/intrusive -The relationship between EE and patient outcome was different by patient ethnicity: An interaction between EE and ethnicity on the DV (patient outcomes)

identifying interactions

•Value of the DV at each level of the interaction -If present, there will be different values at levels of one IV by the level of another IV -If not present, there will be no difference in values in levels of one IV by another IV -Confirm with statistics