Psych 144 Exam

Science

Method of error detection

Systematic Empiricism

learning about the world through careful observation

Empirical Questions

can be presented in such a way that they can be presented as falsifiable

Public Knowledge

allows for other people to test it in new ways

Why is Public Knowledge essential?

• science is a social process-our current scientific knowledge is based on the different studies conducted by many different researchers

• publication allows science to be self-correcting

Skepticism

open but slow to adopt new ideas

Tolerance of Uncertainty

something we don’t really know-no answer-shouldn’t try to force an answer/explanation

Methodological Naturalism

that there is a natural explanation for what you’re seeing

Naturalism

try to explain the world from a natural phenomenon

Pseudoscience

lack one or more of the 3 features of science

What are the 3 features of science ?

• systematic empiricism

• empirical questions

• public knowledge

Why is Pseudoscience harmful?

• harmful to individuals

• harmful to the field of psychology

Doctor of Philosophy

the highest degree in most academic fields, including psychology

Basic Research

scientific research that is concluded primarily for the sake of learning

Applied Research

scientific research that is concluded primarily to solve some practical problem

Validity

basically assessing if the measure actually measures what it claims to

Face validity

simplest form of validity, it is simply whether the measure seems to measure the target construct

Content Validity

how well a measure covers all aspects of the conceptual definition of the construct

Criterion Validity

high when a measure correlated with some other criterion (hence the name) that it logically should (particularly when that criterion is highly reputable)

Discriminant Validity

lack of correlation with variables measuring theoretically distinct concepts/constructs

Construct Validity

how well a measure measures what it claims to

Reliability

produces consistent results under consistent conditions and is consistent with itself

Test-Restest Reliability

simple way to assess that a test produces similar results on repeated administration. Is it consistent over time?

Parallel Forms Reliability

ways to assess the consistency of different versions of a measure

Inter-rater Reliability

multiple judges rate a variable

Internal Consistency

basically a measure of how consistent a measure is with itself

Split-half Reliability (internal consistency)

asses if half the responses on a measure are consistent with the other half

Cronbach’s Alpha

a measure of internal consistency that is far superior to split-half because it essentially measures the mean of all possible split-halves

Measurement

The assignment of values using a set rule or procedure to represent some characteristic of interest

Psychometrics

when measurement is related to psychology

Variables

a quantity or quality that varies across people or situations

Discrete variables

clear separation from each value and numerical values can only be whole integers

Continuous

lack of clear separation and values need not be whole integers

Constructs

Complex variables that cannot be directly observed

Ex: intelligence

Conceptual definition

the behaviors and processes that compose a construct

• what the construct is

Operational definition

how we define and measure a variable/construct

• how we are measuring it for a specific study

self-report measures

use subjective responses from a participant

Behavioral measures

use some aspect of observational behavior

physiological measures

use physiological processes

Ex: stress - cortisol as a measure of stress

Nominal scale

(lowest level) variables are defined by their labels and have no numeric value

Ordinal scale

variables are defined by the order they appear in, but the size of the differences is not clear

Ex: allows rank ordering of responses like agree or disagree

Interval scale

a numeric scale where variables are defined by order, but also the size of the distance is known

Ratio scale

same as interval, but has a value for absolute zero (though the zero can be theoretical)

Descriptive statistics

used to organize and describe the characteristics of a collection of data

Measure of Central Tendency

the central and most typical value, also called the average score. It is the single score that best represents the distribution as a whole and most scores in a distribution tend to cluster around it

Variance

how much do scored tend to differ from each other? What is the overall spread or dispersion of the scores?

Central Tendency

is a measure that reflects the typical score in a distribution

• sometimes referred to as the average

Mean

also called the arithmetic mean, it is the calculated average of a numerical distribution

• the sum of the scores divided by the number of scores

Median

it is the middle score of a numerical distribution

• great for distributions that have outliers

Mode

the most common score

Symmetrical (normal)

the left and right halves are mirror images of each other

Negatively Skewed

the peak shifts toward the upper end of its range and has a relatively long negative tail

Positively skewed

most cluster on the lower end of its range and has a relatively long positive tail

Standard Deviation

the average distance between the scores and the mean

When does a statistical relationship exist?

when the average value on one variable differs systematically across the levels of another variable

Visualizing relationships: Groups and Conditions Effect size-cohen’s d

measures the strength of the difference between two groups or conditions

• differences between means

• .20 is a weak relationship

• .50 is a medium strength relationship

• .80 is a strong relationship

Correlation between quantities: Two quantitative variables Pearsons r

often assessed by calculating the correlation between the two variables

• simply assessing the association between them

• 1 is a perfect positive relationship

• -1 is a perfect negative relationship

• 0 represents no relationship between the variables

• can’t go above 1

• .10 is a small relationship

• .30 is a medium relationship

• .50 is a strong relationship

Correlational Research

research focused on examining correlations between variables without any manipulation of those variables

• simply measuring them and seeing if there’s some kind of relationship

Research Surveys (correlational research)

constructed to measure variables

Naturalistic Observations

observing behaviors in natural settings

Archival Data

use data already stored

Can correlation show causation?

No, it cannot show it by itself because the relationship could be a coincidence, the direction of the relationship cannot be known without additional evidence (directionality), other variables could account for changes in one or more of the variables (third variable)

Directionality

when it is unclear which variable could be the cause and which could be the effect

Third Variable

or extraneous variables - outside the focus of the study

Confounding Variable

this is an extraneous variable that could explain the relationship observed

Experimental Research

an experiment is a specific type of study that involves attempts to limit the number of variables so that specific variables of interest can be observed in relative isolation

• limit the number of extraneous variables

Independent Variable

(Manipulated variable) variable that researcher controls and directly changes - cause the changes

Dependent Variable

outcome or response variable - the response due to manipulated variable

Condition

one for each distinct level of the independent variable

Treatment

a condition or stage of the study that has the independent variable present or applied

Control

a condition that either receives nothing or a placebo

• experiment does not have to have a control condition

Random Assignment

every member of the selected sample has an equal chance to be included in a condition

Counterbalancing

testing different participants in different orders

• a requirement for a true within-subjects experiment

Internal Validity

can we reasonably assume causality from the study?

• tends to be weaker (lower) for a correlational study because of issues involving other variables

• tends to be stronger (high) for experiments because there is an attempt to control for other variables

External Validity

are the results applicable to other settings?

• if researcher controls for extraneous variables, then the setting could be artificial and not reflect how a person or thing would react in other situations

What key groups are influence by research ethics?

• participants

• scientific community - less trustworthy if not ethical

• society - if society is less trusting of science it is harder to make arguments

How is privacy respected?

privacy is respected by proper application of confidentiality and sometimes anonymity

Institutional Review Board

are groups that review proposals for research studies

What is informed consent?

participants should be informed of

• the purpose of the research

• possible ricks and benefits to the participant and society

• length of time expected to take

• procedures involved in the study

• person to contact regarding questions

• their right to refuse participation at any time

• if deception is used, the participant needs to be informed as soon as possible as part of a debriefing

• note that a briefing is only technically required when deception is used or the study elements need elaborating, though they can be used after any study

Phenomenon

an observable fact or occurance that emerges through empirical observation

Dunning-Kruger Effect

overestimates their abilities

Blindsight

a person who is blind but seems to be aware of their enviornment

Bystander Effect

the idea that the more people around, the less likely someone will help or slower to respond

Hypothesis

testable statement/prediction (before you gather your data)

Theory

a coherent explanation for a phenomenon or phenomena

Perspective

a broad outlook that can explain a large number of phenomena

Model

a specific explanation of specific phenomena , they are often derived from and/or are included in theories

Formality

extent that relationships and components are specified

Scope

the number and diversity of phenomena they apply to/explain

Theoretical Approach

types of theoretical ideas they are attached to

Functional Theories

explain phenomena in terms of function or purpose (commonly in evolutionary psych)

Mechanical Theories

explain phenomena specifically in terms of how they work

Organization

organize phenomena in ways that make them easier to understand and think about

Prediction

make predictions about what will happen in new situations

Dunning-Kruger Effect

people of low competence lack the metacognitive ability to properly evaluate their actual level of competence

• if all you know is this little bit, it’s easier to over estimate

Blindsight

the lateral geniculate nucleus of the thalamus may be able to process limited visual data (enough for limited response and unconscious awareness)

Bystander Effect

Stimulus Overload - the information coming into the brain is too much for it to process

Phenomena

observations

What does a true experiment require?

• manipulation of an independent variable

• random assignment to conditions or order of conditions

Types of control conditions?

• no treatment control

• waitlist control

• placebo control