307 - Test 1

on exams - pick things out of the book you learned that was not talked about in class

IMPORTANT

how the data, a theory, and hypothesis work together

form a triangle with data on the bottom left, then it goes up to theory, then it goes down to hypothesis and then back to the data. see 9/30 of notes

looks and acts like a science, but it's not. ex astrology, palm reading, analyzing handwriting, phrenology.. can't test results with science. rely on logical positivism and anecdotal evidence

pseudo science

objective (unbiased) empirical observations of phenomena that are made to occur in a strictly controlled situation in which one or more factors are varied and the others are kept constant

psychological experiment

any data besides numbers (videos, interviews)

qualitative

proving yourself right

logical positivism

what are the 4 ways we assess reliability?

test/retest equivalent forms internal reliability interrater reliability

EXTRA TEXT - observations are recorded during preselected time intervals

time interval sampling

EXTRA TEXT - observations are recorded every time a particular event occurs

event sampling

What are the two approaches to research?

experimental (manipulating. looking at cause and effect) descriptive (describe what's happening, but you don't know why)

the strongest evidence for causality comes from which of the following research methods?

experimental research

why something happens

explanation

outside variable (other than the Iv) that may have an effect on the DV. can cause variance within the scores

extraneous variable

anything other than the IV that can change the DV

extraneous variables (***)

having more than 1 independent variable

factorial design

what are some experimental settings?

field (the real world), lab, online

educated guess. used to test a theory. more specific

hypothesis

making a prediction and comparing it to actual data - proving yourself wrong is sometimes best

hypothesis testing

proving yourself wrong - provides more valuable information

poppers position of falsification

who youre trying to draw conclusions about

population

what will happen - results of something else

prediction

we remember the first and last thing the best

primacy and recentcy

you want to do something new; not replicate other studies for this reason

publication

in what kind of research are words and pictures common forms of data?

qualitative

any data that's just numbers

quantitative

scale of measurement that has an absolute zero. for example you can have a time of 0.0 or a distance of 0.0

ratio

making generalizations usually on memories - study on sleep words. you observed words by listening problem was 7 people wrote down sleep even though it wasn't on the list

rationalization

something is dependable/consistent/stable

reliability

what are the two things you need in order to have a good measurement?

reliability and validity

repeating a study and getting the same results. time, population sample, participants, 5% chance could all affect results

replication

represents the population only on a smaller scale

representative sample

part of a population - actual participants

sample

systematic empiricisms - making assumptions/observations as unbiased as possible. (knowledge based on observation)

science

paying attention to one thing while ignoring another thing

selective attention

what is the 1 advantage of an experimental approach?

shows cause and effect

variable

something that changes

compares means of 2 groups

t test

being persistent/stubborn - ex superstitions; like not washing socks before winning a game. also like a habit. done by chance reinforcement

tennacity

based on research already done (pattern) - try to explain a collection of data. more general

theory

occurs when observed relationship between two variables is actually due to a confounding extraneous variable

third variable problem

EXTRA TEXT - use of multiple data sources, research methods, investigators, and or theories/ perspectives to cross check and corroborate research data and conclusions

triangulation

something measures what its supposed to measure. accurate and correct information

validity

how spread out the scores are

variance

continuous variables

(quantitative) variabes that are numbers

general to specific

deduction

categorical variables

they're categories (ex majors)

characteristics of the scientific method

1. control 2. operationalism 3. replication

what is empiricism?

acquiring knowledge through experience

based on the number of people (sample size)

degrees of freedom

when would you use a t test?

when you're comparing the averages between two groups

how much the participants in the group vary

within groups variance (error variance)

disadvantages to experimental approach

- some things you can't manipulate cause its unethical or impossible - experiments are artificial cause of control - inadequate all by itself - can't just look to experiments for all answers to questions

required conditions to show cause and effect

1. cause must be related to the effect (statistics) 2. cause must precede the effect. cause must happen 1st 3. there must be no other explanation for that effect (keeping control) - eliminating confounding variables

if a analysis gets .065 is it significant?

No because it has to be less than .05. So .048 would be significant

a mediating variable is

an intervening variable

people believe more things from a person in an authority role rather than someone who is not

authority

how different the groups are from each other

between groups variance

what are the 3 pieces of information to figure out if a statistic is significant?

between groups variance (means) within groups variance (standard deviation) sample size (degrees of freedom)

two types of variables

categorical and continuous

a variable that varies in type or kind is called a

categorical variable

manipulating A changes B

causal description

explaining why A changes B

causal explanation

something that makes something else exist; not something that makes something change

causation

what does experimental research show

cause and effect

taking qualitative research and turning it into numbers. important to be able to analyze and get significant results - numbers aren't biased

coding

a variable besides the IV that does change the DV

confounding variable (***)

something that has different effects on each of the manipulated groups. anything other than the IV that does have an effect on the DV. it can make it look like you have results when you don't. eliminates your ability to tell cause and effect

confounding variables

the extent to which from operational definition captures the concept that you wanted to. did it measure what you wanted it to? overarching umbrella term

construct validity

if you know that something causes something else, you can control for the future. applying what you found

control

keeping things the same in an experiment

control

part of the experiment that makes sure your extraneous variables don't become confounding variables

control

what does random assignment help you achieve in an experiment?

control

EXTRA TEXT - what is the role of the scientist in psychological research?

curiosity - discovering something new patience objectivity - eliminate personal bias change

something that changes depending on the independent variable

dependent variable

describe not explain

description

psych objectives

description, explanation, prediction, control

fight of flight response

different than what it was in the past. our ancestors used it to get away from physical short term stressors. now its more long term emotional stress

the difference between what would have happened and what did happen (in an experiment look between experimental group and the control)

effect

must follow a set of rules. must have observed or done data

imperical research

the variable that is presumed to cause a change in another variable is called a

independent variable

specific to general

induction

are your scores on the individual questions about the same. you should get a consistent score not do amazing on the whole test and then get a portion of it wrong

internal reliability

between ratter (people recording) reliability. have at least 2 people watching at the same time - good if they agree on measurements

interrater reliability

scale of measurement that has the same amount of difference between each one

interval

when you feel like something will happen and then it does. just by chance but you feel like its real. our body picks up patterns (gut feelings) before our mind registers it

intuition (helpful in forming hypothesis)

what are the 4 methods of acquiring knowledge?

intuition (knowledge not based on facts) authority (accepting knowledge based on person of authority) rationalism (using reasoning to come to a conclusion) empiricism (knowledge through experience)

the more you've invested in something, the more you're going to like it

investment theory

item total correlation

look at each question to see if they're right or wrong

a link in the middle of 2 things - ex why might two things relate to one another

mediator

combining old studies and analyzing them. quantitative technique for describing the relationship between variables across multiple research studies

meta analysis

merely being exposed to something will make you like it more because its more familiar

mirror exposure effect

changing the direction of something. ex the exception to a rule

moderator

can a measurement be valid without being reliable?

no but it can be reliable but not valid

scale of measurement that's names and categories

nominal

what are the 4 scales of measurement?

nominal, ordinal, interval, ratio

is emperical data scientific?

not unless it follows a set of rules

how you represent or measure a concept

operationalism

science must be specific and precise and each concept must be defined by the steps or operations used to measure them

operationalism

scale of measurement that's order but you don't know how much space is between two things

ordinal

EXTRA TEXT - type of research in which a researcher hypothesizes a theoretical causal model and then empirically tests the model

path analysis