PSY 220 EXAM ONE

conducts studies and create scientific knowledge

• psychologist designs an experiment examining whether sleep affects memory

research producers

read, understand and evaluate research conduct by others

• therapist reads research to decide which treatment methods are supported by evidence

research consumers

using systemic observations and evidence to answer questions rather than relying only on intuition, personal experience or authority

• instead of assuming that studying with music improves grades, researchers collect data comparing students who study with and without music

scientific empiricism

theory - research question - hypothesis - collect data - evaluate results - refine/support theory - new research

• theory explains how or why something happens

• researchers derive hypotheses from it and collect data to see whether the predictions are supported

Theory-Data cycle

theory must make predictions that could potentially be shown to be wrong

• there has to be possible evidence that would contradict it

  • Ex: Getting at least 8 hours of sleep improves memory-test performance compared with getting 4 hours - this can be tested and potentially contradicted

falsifiable

when multiple theories explain the evidence equally well, scientists generally prefer the simpler explanation that requires fewer unnecessary assumptions

• simplest adequate explanation

parsimonious

scientific studies generally do not prove a theory is true

• researchers say:

  • “the evidence supports the hypothesis”

  • the results are consistent with the theory

  • the evidence does not support the hypothsis

Doesn’t “prove” anything

Research designed to increase knowledge and understand how or why something happens, without necessarily trying to solve an immediate real-world problem

• ex: studying how stress affects attention and memory

Basic research

research that takes findings from basic research and investigates how they can be used in practical situations

• using what researchers know about stress and memory to develop a possible stress-reduction technique for employees

translational research

research conducted to solve a specific, real-world problem or improve an existing situation

• testing whether a stress-reduction program actually lowers stress among retail employees

applied research

• done to show an effect is reliable

  • if an effect is reliably demonstrated across many studies, it is considered reliable

replication

only 36% of findings are replicated

• social priming and precognition research studies

replication crisis in psychology

specific to general

• start with observations/data and use them to develop broader explanations or theory

• notice a pattern and might ask what it means

  • ex: you observe that several retail employees report feeling stressed after dealing with rude customers - so you propose customer mistreatment may increase employee stress

Inductive reasoning

general to specific

• start with an existing theory or general idea and make a specific prediction that you can test

  • ex: theory says customer mistreatment increases stress - so you predict that employees who experience more customer mistreatment will report higher perceived stress

deductive reasoning

tendency to judge how common or likely something is based on how easily examples come to mind

• ex: after seeing several news stories about airplane crashes, someone might believe flying is extremely dangerous because those examples are easy to remember

availability heuristic

an organized explanation of observations that can be used to make testable predictions

scientific theory

fits the available evidence and observations

• research findings generally support the theory’s predictions

accuracy

doesn’t contradict itself and is compatible with established evidence

• theory give logically consistent explanations across studies

consistency

explains a wide range of observations, not just one specific finding

• a theory of stress explains stress responses in several different situations

scope

use the simplest explanation possible without unnecessary assumptions

• called parsimony

• if two theories explain the same evidence, prefer the one requiring fewer assumptions

simplicity

generates new research questions, hypotheses and discoveries

• a theory about stress leads researchers to test how sleep, work, or social interaction affect stress

fruitfulness

method researchers use to determine whether their results are unlikely to have occurred by chance if the null hypothesis were true

Null hypothesis significance testing

• predicts that there is a relationship, different or effect

• usually reflects what the researcher expects to find

  • ex: employees who experience more rude customers will report higher stress levels

research hypothesis

• predicts there is no relationship, no difference or no effect

• researchers use statistical tests to determine whether there is enough evidence to reject the null hypothesis

  • there is no relationship between customer mistreatment and employee stress

null hypothesis

you conclude that an effect exists when it actually doesn’t

• You conclude rude customers are related to employee stress, but in reality there is no relationship

• “I found something that isn’t really there”

Type 1 error

you conclude there isn’t enough evidence for an effect when an effect actually exists

• yur study finds no significant relationship between rude customers and stress, but customer mistreatment actually is related to stress

Type 2 error

statistical significance is about probability not proof

p < 0.5

• means that assuming the null hypothesis is true, the probability of obtaining results this extreme or more extreme is less than 5%

probability v proof

reject that H0

p < 0.5

fail to reject H0

p > 0.5

summary of the entire study: purpose, methods, major results and conclusions

• what is this study about?

abstract

background information, previous research/literature, theory, research question and hypothesis

• explains why the study was conducted

  • "why did they do the study

introduction

exactly how the study was conducted: participants, materials/measures, variables, design and procedure

• how did they do it?

methods

statistical findings and analysis

• includes tables, graphs, p-values, means, correlations, etc

  • what did they find

results

researchers interpret the results, explain what they mean, connect them to previous research, discuss limits and suggest future research

• what do the findings mean

discussion

complete citations for the sources mentioned throughout the article

• what sources did they use

references

summarize existing literature

• synthesize many studies into a larger theory

review articles

analyze data from multiple scientific articles

meta-analysis

introduce new ideas, reflect, challenge dogmas

• less than 5k words

commentary/opinion

report new research findings

empirical research

decide whether a research article should be published in an academic journal

• editor checks for:

  • reviews if it fits the journal

  • decides whether to send it to peer reviewers

  • consider the reviewer’s feedback

  • makes the publication decision

journal editors

article will be published

• sometimes after minor changes

accept

article will not be published

reject

researchers need to make changes based on feedback and submit article again

• does not guarantee acceptance

revise and resubmit

other experts in the same area evaluate the research process before publication

• process: article is submitted - editor reviews it - peer reviewers evaluate it - reviewers provide feedback - editior makes decision

peer review process

• research methods

• statistical analysis

• whether conclusions match the results

• importance of the research

• problems or limitations

• quality and clarity of the article

purpose is to help maintain the quality and credibility of scientific research

peer reviewers look at

measures a researcher’s publication impact using

• number of publications

• number of citations

a researcher has an H-index of H when they have H papers that have been cited at least H times

H index

Step 1: Put citation counts from highest → lowest.

Step 2: Number the papers 1, 2, 3, 4, 5...

Step 3: Compare:

Citations ≥ Paper Number

Step 4: Find the last paper where that statement is true.

That paper number = H-index.

calculating H index

condition is manipulated or selected by the experimenter to determine effects on behavior

independent variable

different conditions or categories of the independent variable

• IV = customer attitude

  • Level 1: rude customer

  • Level 2: polite customer

levels of independent variable

independent variable that the researcher does NOT manipulate because it is a characterisitc participants already have

• ex:

  • age

  • biological sex

  • personality traits

  • prior work experience

subject variable

researcher creates/assigns the conditions

manipulated IV

outcome that is measured to determine whether it was affected by the IV

dependent variable

explains exactly how a variable will be measured or manipulated in a study

• turns an abstract or vague idea into something that can be observed, measured and quantified

operational definition

an abstract concept that cannot necessarily be directly observed

• ex:

  • stress

  • happiness

  • intelligence

  • anxiety

construct

places people or observations into groups or categories

• ex:

  • rude v polite customer

  • employed v unemployed

  • freshman, sophomore, junior, senior

categorical

represents an amount or numerical quantity

• ex:

  • age

  • stress score

  • number of rude customers

quantitative variable

has equal distance between values

• has no true zero

  • The difference between 60F and 70F is the same amount as the difference between 70F to 80F

interval

equal distances between values

• has a true zero

  • age, weight, height, reaction time

ratio

numerals represent a rank order

distance between each number may not be equal

• ex:

  • 1st place

  • 2nd place

  • 3rd place

ordinal

third variable that you are not interested in studying, but it changes along with the independent variable

• creates a problem because you can’t tell whether the IV or the confound caused the change in the DV

confound variables

measures consistency between different observers or raters

• two or more researchers independently observe or evaluate the same behavior

• their ratings should be similar

interrater

measures consistency over time

• same participants take the same test on two different occasions

• researchers compare the score to see whether they are similar

  • students take a personality test on Monday and again two weeks later. If their scores are similar, the test has a high retest reliability

test-retest reliability

measures how consistently different items on the same test construct

• used for questionnaires and tests with multiple questions

• items intend to measure the same construct should produce related responses

  • a stress questionnaire contains 10 questions about perceived stress. If participants who score high on one stress-related question generally score high on the other related questions, the questionnaire demonstrates internal consistency

internal consistency reliability

does the measurement actually measure the construct it is intended to measure?

• A researcher develops a questionnaire about stress. They must determine whether the questions actually measure stress rather than depression of anxiety

construct validity

how confidently researchers can conclude that the independent variable caused changes in the dependent variable, rather than other factor

• high internal validity means there are few plausible alternative explanations for the results

• confounding variables threaten internal validity

• random assignment and controlling extraneous variable can help increase internal validity

internal validity

how well the findings of a study can be generalized to other people, settings, and situations

• can the results apply to people outside the study

• would similar results occur in different environments

• would the findings hold under real-world conditions

external validity

purpose: investigate how far people would go in obeying an authority figure, even when instructued to harm another person

• participants were instructed to administer increasingly powerful electric shocks to another person whenever they answered a question incorrectly

• person receiving the shocks was an actor and no real shocks were administered

• participants believed the shocks were real

• experimenter instructed participants to continue even when they expressed discomfort

milgram’s obedience to authority studies summary

• deception: participants were misled into believing they were administering real electric shocks

• Psychological harm: many experienced significant stress, anxiety and emotional distress

• Right to withdrawal: experimenter’s instructions pressured participants to continue, making withdrawal difficult

milgram’s obedience to authority studies ethical issues

purpose: investigate how assigned social roles and prison environments influence behavior

• male college students were randomly assigned to be either prisoners or guards in a stimulated prison

• some guard engaged in humiliating and abusive behavior toward prisoners

• some prisoners experienced considerable emotional distress

• study was scheduled to last two weeks, but ended after 6 days

Stanford Prison Experiment Summary

• Psychological harm: participants experienced humiliation and emotional distress

• Right to withdrawal: some prisoners reportedly encountered difficulties leaving the experiment

• Researcher bias: Zimbardo acted as both the principal investigator and prison superintendent, creating a conflict between his research responsibilities and his role in simulation

• Inadequate protection: researchers did not adequately intervene when participants experienced harm

Stanford Prison Experiment ethic problems

purpose: observe the natural progression of untreated syphilis in Black men

• U.S. Public Health Service recruited about 600 Black men in Alabama, including 399 with syphilis and 201 without it

• participants were not correctly informed about the true purpose of the study or their diagnoses

• researchers continued observing participants without providing appropriate treatment, even after penicillin became the recommended treatment for syphilis

• The study continued for 40 years, ending in 1972

tuskegee physillis study summary

• lack of informed consent: participants were deceived about the purpose of the research and their medical conditions

• Withholding treatment: researchers denied participants appropriate treatment, exposing them to preventable harm

• Racial exploitation: study exploited Black men from a disadvantage community

• Violation of beneficence: researchers failed to protect participants from serious and avoidable harm

tuskegee physillis study ethical issues

3 established major ethical principles to protect human principles in research

• developed in response to unethical research practices, including the Tuskegee Syphilis Study

Belmont Report 1979

researchers must respect participants’ autonomy, meaning their right to make their own decisions about participating in research

• participation must be voluntary

• participants must be able to withdrawal without penalty

• Researchers must obtain informed consent

• individuals with limited autonomy require additional protected

respect for persons

groups that may have limited autonomy or face increased risks of coercion or exploitation

• ex: this includes children, prisoners, and individuals whose decision-making capacity is impaired

voluntary populations

researchers must protect participants from harm while maximizing possible benefits

• Do not harm participants

• maximize potential benefits and minimize potential harms

beneficence

the benefits and burdens of research should be distributed fairly

• researchers must select participants fairly rather than targeting certain populations simply because they’re accessible, disadvantage or vulnerable

  • researchers should not recruit low-income individuals for a risky medical study while reserving its potential benefits exclusively for wealthier populations

justice

examines:

• purpose and procedures of study

• potential risks and benefits

• infomed consent procedures

• protection of vulnerable populations

• privacy and confidentiality

• whether participant selection is fair

IRB review process

research that qualifies for specific regulatory exemption categories, often involving low risk activities

• ex:

  • anonymous surverys about study habits

  • certain educational observations

  • some research involving existing, nonidentifiable data

exempt research

probability and magnitude of harm or discomfort are no greater than those ordinarily encountered in daily life or during routine physical or psychological examinations

• ex:

  • noninvasive EEG recordings

  • certain blood samples

  • some interviews involving mildly sensitive topics

minimal risk research

research involving greater than minimal risk generally requires review by the convened IRB committee

• ex:

  • experimental procedures with significant physical risks

  • studies involving substantial psychological distress

  • research involving potentially dangerous interventions

greater than minimal risk research

protects the rights and welfare of human research participants

• focuses on informed consent, privacy, confidentiality, risks, benefits and fair participation selection

IRB

instituional animal care and use committee

• oversees research involving animals to ensure their humane treatment and appropriate care

• evaluate animal housing, veterinary care, pain management, scientific justification and alternatives to animal use

IACUC

replacement: use alternatives to animals when scientifically appropriate

reduction: use the fewest animals necessary to obtain scientifically valid results

refinement: modify procedures to minimize animal pain, suffering and distress

Three R’s of Animal Research

food and consumer behavior researchers at Cornell. His research examined how environmental factors, like plate sizes and food presentation and influence eating behavior

• there was academic misconduct involving misreported data, problematic statistical techniques, poor recordkeeping and inappropriate authorship

• his work had p-hacking

brian wansink purpose

published a blog post describing how researchers in his lab reanalyzed a dataset to produce additional publications

• other scientists began examining his published studies and identified statistical errors, inconsistencies and questionable research practices

• papers were redacted and he was resigned in 2018

brian wansink caught

known for researching dishonesty, co-authored in a study suggesting that people report information more honestly when they sign an honesty declaration at the beginning of a form rather than the end

• researchers discovered that data from an insurance company experiment included in the paper had been fabricated

• the paper was redacted

• ariely denied fabricating the data and said it had been supplied by the insurance company

dan ariely purpose

independent researchers associated with the blog Data Colada examined the publicly available data and identified unusual statistical patterns, including an implausible distribution or report vehicle mileage

dan ariely caught

• fabricated and manipulated data

• he created datasets for experiments that had never taken place

  • also fabricateddata to colleagues and doctoral students who did not know the data was false, led to 55 fraud publications

diederik stapel purpose

three junior researchers became suspicious because his results appeared unusually consistent with his hypothesis

• they investigated his datasets, identified irregularities and reported their concerns to university officials in 2011

• stapel admitted to fabricating data and was dismissed from Tilburg university

diederik stapel caught

researcher develop or change their hypothesis after examining their results but present those hypotheses as through they were established before the study

• an unexpected discovery appear to confirm a prediction made in advance, potentially misleading readers about the strength of evidence

HARKing

researcher manipulate their data collection or statistical analyses to obtain statistically significant results

p-hacking

checking results repeatedly and stopping data collection once a significant result appears

• researcher plans to recruit 10 participants but stops at 40 because p=0.04

stopping early

selectively including or excluding observations to obtain desired results

• removing participants whose scores do not support the hypothesis without a valid reason

cherry-picking

testing many relationships without accounting for the increased likelihood of false positives

• testing 20 different relationships and reporting only the significant ones

multiple comparisons

trying different statistical analyses until one produces a significant result

• trying several tests and reporting only the one yielding p = 0.03

shopping statistical tests

intentionally concealing observations, results or relevant information

• researcher conducts 5 experiments but publishes only the two that supports the researcher’s prediction

withholding data

a study was published involving 42 participants

• reported adopting expansive, powerful looking body postures increased feelings of power and risk-taking, raised testosterone and lowered cortisol

• larger replication study found increased self-reporting feelings of power but did not reproduce the hormonal or behavioral effects

• methodological problems were acknowledged in the original study, including checking results during data collection, a small sample and reporting a more favorable statistical tests

power pose studying

• practices that can distort findings or exaggerate evidence

• examples include undisclosed p-hacking selective reporting and HARKing

• can occur without an intention to receive

questionable research practices

• deliberate misconduct involving fabrication, falsification or plagiarism

• examples include inventing participants or deliberately altering recorded results

• involves intentional knowing or reckless misconduct under applicable research standard

research fraud