Psychology Chapter 2

Variable

any characteristic/factor that can vary

What is a factor? (in relation to factor analysis)

"label" for the cluster aka the group of highly correlated variables

What is internal validity?

Degree to which the experiment supports clear, causal conclusions. If the experiment is designed carefully=high internal validity

What is naturalistic observation, how is it good/bad?

Naturalistic Observation: observe behaviour in natural setting Advantages: provides rich descriptions of behaviour in real world Disadvantages: not good for cause & effect, measurement (observer) bias, present of researcher may effect behaviour

Is it correct to say you prove an experimental hypothesis or reject the null hypothesis?

Never prove the experimental hypothesis, instead say we reject the null hypothesis (ex. The difference is not due to chance= reject null hypothesis)

What is a placebo and the placebo effect? (threat to internal validity)

Placebo: substance with no pharmacological effect Placebo effect: peoples health can sometimes improve simply because they expect it to

What is an independant variable?

manipulated by experimenter, varies between people or across time

3 scientific attitudes

curiousity (want to know why), skepticism (show me the evidence), open mindedness (about evidence and conclusions)

Operational Definition

defines a variable in terms of specific procedures used to measure it Answers question: how will we measure it?

What is external validity?

degree to which the results can be generalized to other populations, settings and conditions

What is correlation?

degree to which the variance in one variable can be accounted for by the variance in another Asking: how well can one variable be predicted by another? NOT CAUSE

what is descriptive research?

described how people and animals behave

What is statistical significance?

difference is statistically significant if not due to chance (think it is a "real" effect) Usually use 5% or less as the cut off

What is stratified random sampling?

divide pop. into subgroups, make sure sample has same proportions as the population

What is a cluster? (in relation to factor analysis)

groups of variables that are highly correlated

theory

set of formal statements, explains how and why events are related

What is frequency distribution? (stats)

shows how many people got what score (ex. How well you did in relation to the class)

What is a correlation matrix? (in relation to factor analysis)

shows the correlations between all variables

What does a positive or negative, and absolute number value of the Pearson Product-Moment correlation coefficient mean?

+: positive correlation -: negative correlation big number: strong correlation, small average distances between line of best fit and points small number: weaker correlation, larger average distance between LOBF and points

What are the 3 characterists of an experiment?

1. Manipulate one or more variable: a. Independent variable 2. Measure the effect on other variables: a.Dependent variable 3. Control extraneous factors

What are the steps in the scientific process?

1. identify question of interest 2. gather info/ form hypothesis 3. test hypothesis through research 4. analyze data, draw conclusions, report findings 5. build body of knowledge

What is inferential statistics?

Allow us to make conclusions about the population from which the sample was drawn (unable to test whole population), tells us how confident we can be in our conclusions

What is meta analysis?

Analysis of all the studies, treats each study as if it was one data point, combines results into one

What is the Pearson Product-Moment correlation coefficient?

Called "r", number ranging from -1.00 to +1.00 that summarizes direction and strength between two variables

What are some threats to internal validity?

Confounding variables, placebo effect, experimenter expectancy

What are measures of variability and 3 ways of measuring? (stats)

Describe the spread of scores Range: highest minus lowest score Variance: first calculate deviation scores (how far each individual score is from the mean), then square the deviation scores, calculate average of squared values Standard deviation: square root of variance

What are measures of central tendency and 3 ways of measuring? (stats)

Describe the typical score Mean: average, can be affected by extremes Mode: most common, can be effected by extremes Median: half scores above and half below, can miss important data

Explain the Between Subjects design of an experiment. What is the problem/solution?

Different groups of participants are assigned to each level of the IV Problem: groups need to be equivalent Solution: random assignment, equal chance balances out any differences between the groups does not eliminate differences Alternative to random assignment: matching, trying to control every variable **Random assignment better than matching b/c tends to equate all variables rather than just those thought of **

Why does correlation not imply causation? (2 problems)

Directionality problem: uncertain which variable changes which Third variable problem: two variables may be related but caused by a third variable (leads to spurious (artificial) correlations)

Explain the within subjects design of an experiment. What is the problem/solution?

Each participant is exposed to all the levels of the IV, aka repeated measure design Problem: order of the conditions can make a difference Solution: counterbalancing, the order of the conditions is varied between people, randomly assign the order

Why do correlations if they dont tell us about causation?

Establish real world associations between variables, some variables cant be measured (ex. religion) but can measure and see how they relate, use correlations to make predictions about behaviour

Compare an experimental hypothesis and null hypothesis

Experimental hypothesis: states there is a real difference between the groups Null hypothesis: states there is no difference between the groups, even if we find a difference, the null hypothesis says that it was a fluke- due to chance

What are the two approaches to understanding? How are they good and bad?

Hindsight: Relies on the explanations after the fact Good: can use to create testable hypothesis Bad: past events can be explained in many ways Through scientific method: relies on prediction, control, theory good: satisfies curiousity, builds knowledge, generates principles Less good: prediction does not necessarily = understanding

What is validity?

How well an experiment actually tests what it is supposed to test

What is a case study, how is it good/bad?

In depth analysis of an individual, or an event Advantages: challenge validity of theories, make new ideas Disadvantages: not good at determining cause & effect, cant generalize (based on 1 person), measurement (observer) bias

What is the "tri council policy"?

In ethics federal agencies have outlines a tri council policy for ethical conduct for research involving humans

tests that measure different types of variables: physiological measures

Includes heart rate, BP, respiration rate etc. Don't always know what these mean ex. What is the link between heart rate and emotion? Is it love? Stress? Excitement

What are some characteristics of a good theory?

Incorporates existing facts/observations meaningfully, is testable, conforms to law of parsimony (simpler is better)

What is incomplete disclosure? (ethics)

Mislead participants about what study is about Sometimes is the only way to obtain natural, spontaneous behaviours, allowed only if no feasible alternative, educational benefits outweigh ethical costs. If this is used participants must be debriefed after

How do you obtain a representative sample?

Random sampling: everyone has equal chance of selection Stratified random sampling: divide pop. into subgroups, make sure sample has same proportions as the population

What is factor analysis?

Reduces a large number of correlations to a smaller number of "clusters", psychologists need to look at cluster and determine underlying factor

What is the experimental method?

Research technique which one or more variables is manipulated to determine if it influences other variables Only method that allows for the conclusion about causes

What is the process of the correlation method?

Researcher measures one variable, measures second, statistically determines if two are related 1. Collect data: difficult to tell if variables are related at this point 2. Make scatterplot 3. Draw straight line of best fit 4.Calculate a Pearson Correlation Coefficient: statistic that tells us how related two variables are

Why are internet discussions a controversy in ethics ?

Researchers analyze online discussions Confidentiality and privacy concerns; informed consent and debriefing participants is difficult

What is survey research, how is it good/bad?

Survey Research: info through questionnaires, surveys, interviews (ex. political polls) need a representative sample Advantages: efficient for collecting lots of data, reveal changes in habit/beliefs over time Disadvantages: cannot determine cause & effect, based on self reports (bias, inaccurate self perceptions etc.)

What is a normal curve? (stats)

Symmetrical, bell shaped, represents theoretical distribution of scores in a population

What is experimenter expectancy? (threat to internal validity)

The experimenters expectations effect the results Solution: researchers interacting with patients are "blind" to hypothesis or to participants conditions Double blind procedure: experimenter and researcher doesn't know what condition the participant is in

What are extraneous factors?

Things that can affect the DV other than the IV, need to make sure these don't affect results, avoid confounding variables (correlates with IV and DV)

What is a confounding variable? (threat to internal validity)

Variables that are intertwined with the IV Prevents us from knowing whether it was the IV or the confounding variable that actually caused the effect ex. Mozart effect

Explain the way to measure variables: Self reports and reports by others

ask people to report own knowledge, behaviours, feelings etc., hinges on honesty

What is random sampling?

everyone has equal chance of selection

Whats a histogram? (stats)

graph of the frequency distribution

Psychological variable

includes self esteem, stress etc

What is a dependant variable?

measured to see if the IV influences it, the "outcome" variable

Habituation

process which organisms eventually ignore the researcher

Explain the way to measure variables: measures of overt behaviour

record directly observable behaviour measurements need to be reliable (specific coding system ensures consistency), and unobtrusive (not obvious it is being measured because people/animals act differently when being watched)

What is a representative sample?

reflects important characteristics of the entire population, not having one leads to very wrong results (ex. 1936 American Election only rep'd rich ppl)

Hypothesis

specific prediction

What is the correlation method?

technique where two or more variables are measured to see if they are related in a systematic way, does not manipulate anything Ex of research questions: are there more hospital emergencies on

Social desirability bias

tendency to respond in a socially acceptable manner, rather than how one truly feels Avoid by: anonymity, question wording

Explain the way to measure variables: psychological measures

tests that measure different types of variables ex. personality test