Midterm 1 Flashcards
Intuition
When you rely on intuition, you accept unquestioningly what your personal judgment or a single story about one person’s experience tells you about the world.
illusory correlation
occurs when we focus on two events that stand out and occur together.
scientific skepticism
Recognizing that our own ideas are just as likely to be as wrong as anyone else’s, and question other people’s pronouncements of truth, regardless of their prestige or authority.
empiricism
knowledge is based on structured, systematic observations
Universalism (1/4 norms)
Scientific observations are systematically structured and evaluated objectively using accepted methods of the discipline. By relying on empiricism in this way, we expect that scientists can conduct research to test any idea, other scientists can disagree, and the research reponed from both sides can be objectively evaluated by others to find truth.
Communality (2/4 norms)
Methods and results are to be shared openly. One major benefit to open reporting is that others can replicate the methods used to check whether they obtain the same results (see Chapter 14 and Collaboration, 2013). Replications help to ensure that effects are not just false positives or random flukes (see Chapter 13). Another major benefit to open reporting is that the results of many studies can be combined in meta-analyses, which are studies that combine results from many studies of the same phenomenon to examine the overall effect (see Chapter 14). No single study provides a perfectly accurate answer to a complex question; a meta-analysis is an important tool in the search for knowledge that relies crucially on communality (see Braver, Thoemmes, & Rosenthal, 2014; Cumming, 2014). Some researchers have begun posting data sets and full study procedures online after a study is published for others to use.
Disinterestedness (3/4 norms)
Scientists are expected to search for observations that will help them make accurate discoveries about the world. They develop theories, argue that existing data support their theories, conduct research to evaluate propositions of their theories, and revise their theories as needed to more accurately account for new data. Scientists should be rewarded for their honest and careful quest for truth, and ideally are not motivated primarily for personal gain.
Organized skepticism (4/4 norms)
All new evidence and theories should be evaluated based on scientific merit, even those that challenge one’s own
work or prior beliefs. Science exists in a free market of ideas in which the best ideas are supported by research, and scientists can build upon the research of others to make further advances. Of all the ideals, organized skepticism is the one that most directly underlies the practice of peer review. Before a study is published in a scientific journal, it must be reviewed by other scientists who have the expertise to carefully evaluate the research and recommend whether the research should be published. This review process, although imperfect, helps to ensure that research with major flaws in theory, methodology, analyses, or conclusions will not become part of the scientific literature.
Falsifiable ideas
it can be either supported or refuted using empirical data
○ If an idea is falsified when it is tested, science is also advanced because this result will spur the development of new and better ideas.
pseudoscience
which uses scientific terms to make claims look compelling, but without using scientific data.
Four general goals of scientific research
(1) to describe behaviour, (2) to predict behaviour, (3) to determine the causes of behaviour, and (4) to understand or explain behaviour.
Criteria for causal claims
- When the cause is present, the effect occurs; when the cause is not present, the effect does not occur - covariation of cause and effect
- There is a temporal order of events in which the cause precedes the effect. This is called temporal precedence
- Nothing other than a causal variable could be responsible for the observed effect. This is called elimination of alternative explanations.
- There should be no other plausible alternative explanation for the relationship.
basic research
attempts to answer fundamental questions about the nature of behaviour.
Applied research
conducted to address practical problems and potential solutions.
Program evaluation research
tests the efficacy of social reforms and innovations that occur in government, education, the criminal justice system, industry, health care, and mental health institutions.
classical conditioning
a neutral stimulus (such as a tone), if paired repeatedly with an unconditioned stimulus (food) that produces a reflex response (salivation), will eventually produce the response when presented alone.
theory
a system of logical ideas that are proposed to explain a particular phenomenon and its relationship to other phenomena
○ theories organize and explain a variety of specific facts or descriptions of behaviour.
○ theories generate new knowledge by focusing our thinking so that we notice new aspects of behaviour.
parsimony
the least complex theory is most desirable, because it is easiest to entirely falsify
abstract
summary of the research report.
introduction
outlines the problem that has been investigated.
method
provides information about exactly how the study was conducted, including any details necessary for the reader to replicate (repeat) the study.
results
presents the findings, which have been based on statistical analyses.
discussion
reviews the current study from various perspectives.
literature review
article using narrative techniques
research hypothesis
- an assertion of one possible state of the phenomenon or relationship under investigation
- statement about something that may or may not be true, is informed by past research or derived from a broader theory, and is waiting for evidence to support or refute it.
prediction
After designing the study, the researcher would translate the more general hypothesis into a specific prediction concerning the outcome of this particular experiment. Predictions are stated in terms of the specific method chosen for the study.
falsifiability
data could show that a hypothesis is false, if in fact it is false.
variable
any event, situation, behaviour, or individual characteristic that can take more than one value (i.e., it varies).
non-experimental method/correlational method
- relationships studied by observing or otherwise measuring the variables of interest. Examine whether variables correlate or vary together, but cannot make statements of causation
- both variables are measured
experimental method
involves direct manipulation and control of variables. The researcher manipulates the first variable of interest and then observes the response.
operational definition
- a definition of the variable in terms of the operations or techniques used to measure or manipulate it in a specific study.
- a description of how a concept will be measured
situational variable
describes characteristics of a situation or environment: the length of words that you read in a book, the credibility of a person who is trying to persuade you, the number of bystanders to an emergency. (can be measured in ANY design, or manipulated in experimental designs. )
Response variable
the responses or behaviours of individuals, such as reaction time, performance on a cognitive task, and degree of helping a victim in an emergency. (are measured in either experimental or non-experimental designs. )
Participant variable
describes a characteristic that individuals bring with them to a study, including cultural background, intelligence, and personality traits such as extraversion.
confounding variables
variables that are intertwined with another variable so that you cannot determine which of the variables is operating in a given situation
correlation coefficient
A numerical index of the strength of relationship between variables
positive linear relationship
increases in the values of one variable are accompanied by increases in the values of the second variable.
negative linear relationship
increases in the values of one variable are accompanied by decreases in the values of the other variable.
curvilinear relationship
increases in the values of one variable are accompanied by both increases and decreases in the values of the other variable (direction of relationship changes at least once)
inverted-U relationship
a curvilinear relationship that increases until a certain point and then decreases
mediating variable
a psychological process that occurs between two variables that helps to explain the relationship between them
random variability/error variability
randomness in events - research is aimed to reduce this by finding relationships between variables
third-variable problem
extraneous variables may be causing an observed relationship.
independent variable
the cause (manipulated)
dependent variable
the effect (observed)
Covariation
in experimental method - when participants in an experimental condition show a different effect relative to participants in a control condition
validity
- “truth” and the degree to which a claim is accurate
Internal validity
- the ability to draw accurate conclusions about causal relationships from our data.
- the degree to which all confounding variables have been controlled and causality can be inferred
- ability to have confidence that the result is because of the manipulated variable
Experimental control
treat participants in all groups in the experiment identically so the only difference between groups is the independent variable.
random assignment
ensures that extraneous variables are just as likely to affect one experimental group as they are to affect the other group, as long as there are enough participants in the study.
field experiment
the independent variable is manipulated in a natural setting.
reliability/precision
the consistency or stability of a measure of behaviour.
true score
the person’s real score on the variable
Measurement error
To the extent that a measure of intelligence is unreliable, it contains measurement error and so cannot provide an accurate indication of an individual’s true intelligence.
Pearson product-moment correlation coefficient.
most common correlation coefficient discussing reliability
- (symbolized as r) can range from 0.00 to +1.00 and 0.00 to - 1.00.
- correlation of 0.00 says two variables are not related at all.
- The closer a correlation is to either +1.00 or - 1.00, the stronger is the relationship.
- positive and negative signs provide info about direction of the relationship.
- When correlation coefficient is positive, there is a positive linear relationship - high scores on one variable are associated with high scores on the second variable.
- A negative linear relationship is when high scores on one variable are associated with low scores on the second variable
Test-retest reliability
assessed by giving many people the same measure twice.
Alternate forms reliability
administering two different forms of the same test to the same people at two points in time.
Internal consistency reliability
assesses how well a certain set of items relate to each other.
Cronbach’s alpha
- a value that is a common indicator of internal consistency
- how well each item correlates with every other item
Interrater reliability
the extent to which raters agree in their observations.
Construct validity
- the adequacy of a variable’s operational definition.
- must be based on relevant behaviours
face validity (construct validity)
the evidence for validity is that the measure appears “on the face of it” to measure what it is supposed to measure.
Content validity (construct validity)
based on comparing the content of the measure with the theoretical definition of the construct
predictive validity (construct validity)
- Using the measure to predict some future behaviour
- The criterion used to support construct validity is some future behaviour.
Concurrent validity (construct validity)
- assessed by research that examines the relationship between the measure and a criterion behaviour at the same time
- having the gold standard being compared to your measurements
- based on an established test (SAT or GPA for intelligence)
Convergent validity (construct validity)
- the extent to which scores on the target measure in question are related to scores on other measures of the same construct or similar constructs.
- comparing to other tests
- eg. new IQ test to correlate with old IQ tests
discriminant validity (construct validity)
- When the measure is not related to variables with which it should not be related
- measure should discriminate between the construct being measured and other unrelated constructs.
reactivity
awareness of being measured changes an individual’s behaviour
reactive measure tells what the person is like when they’re aware of being observed, but it doesn’t tell how the person would behave under natural circumstances.
Nominal scale
no numerical or quantitative properties. Categories or groups just differ from one another (also called categorical variables).(sex, colour)
Ordinal scale
allows us to rank order the levels of the variable being studied. (degrees of burn)
interval scale
- the difference between the numbers on the scale is equal in size.
- No absolute zero to indicate the absence of the variable (temperature in Celcius)
ratio scale
- the difference between the numbers on the scale is equal in size.
- has an absolute zero point that indicates the absence of the variable being measured (age)
Survey research
uses questionnaires and interviews to ask people to provide information about themselves-their attitudes and beliefs, demographics (age, sex, income, marital status, and so on) and other facts, as well as past or intended future behaviours.
panel study/longitudinal design.
When the same people are tracked and surveyed at two or more points in time
response set
the tendency to respond to all questions from a particular perspective rather than to provide answers that are directly related to the questions.
- reduces the usefulness of data
social desirability response set
makes people answer in the most socially acceptable way
Unnecessary Complexity in questions
The questions asked in a survey should be relatively simple.
double-barreled questions
questions that ask two things at once
loaded question
question written to lead people to respond in one way
Negative wording (questions)
Avoid phrasing questions with negatives (“do you agree they should not approve of…”
“yea-saying” or “nay-saying” response set
when you ask several questions about a topic, a respondent may either agree (yea) or disagree (nay) with all the questions.
closed-ended questions
a limited number of response alternatives are given
open-ended questions
respondents are free to answer in any way they like.
Rating scales
these ask people to provide “how much” judgments on any number of dimensions - the amount of agreement, liking, or confidence, for example.
graphic rating scale
requires a mark along a continuous 100 mm line that is anchored with descriptions at each end.
semantic differential scale
a way to measure the meaning that people ascribe to concepts
Non-verbal scale
using images instead of words or numbers (smiley faces vs sad faces)
Questionnaires
in written format, and respondents write or type their answers.
population
a set of people of interest to the researcher.
proper sampling
allows us to use information obtained from the respondents who were sampled to estimate characteristics of the population as a whole.
confidence interval
a range of plausible values for the population value; values outside the confidence interval are implausible
sampling error/margin of error
the error that exists in the estimate because only a sample and not the entire population was measured
external validity
- findings based on a sample can be generalized to the broader population
- ensuring that the sample is highly representative of the population from which it is drawn.
sampling frame
the actual population of people (or clusters) from which a random sample will be drawn
probability sampling
- each member of the population has a specifiable probability of being chosen
- When you wanna make precise statements about a specific population
- used for phenomena that are expected to vary across the population
non-probability sampling
we don’t know the probability of any particular member of the population being chosen
- used for phenomena that are relatively similar across the population so it doesn’t matter who participates
simple random sampling
every member of the population has an equal probability of being selected for the sample.
random sample
When people are randomly selected from a specific population to participate in a study
stratified random sampling
the population is divided into subgroups/strata, and then simple random sampling is used to select sample members from each stratum.
cluster sampling
researcher can identify “clusters” of people and sample from these clusters. After the clusters are chosen, all people in each cluster are included in the sample
non-probability or non-random sampling techniques
A population may be defined, but little effort is expended to ensure that the sample accurately represents the population
convenience sampling/haphazard sampling.
Participants are recruited wherever you can find them.
purposive sampling
purpose is to obtain a sample of people who meet some predetermined criterion
quota sampling
a sample that reflects the numerical composition of various subgroups in the population
selection differences: (have to be eliminated)
The people selected to be in the conditions should not differ in any systematic way. (high income given manipulated and low income given control)
Independent groups design/between-subjects design
different participants are assigned to each level of the independent variable using random assignment.
pretest-posttest design
when they give a test to ensure that participants are equivalent with a certain variable (language, intelligence)
posttest-only design.
when no pretest is given
mortality
dropout factor in experiments
Solomon four-group design
a type of complex experimental design that allows a researcher to assess the impact of the pretest directly.
repeated measures design/within-subjects design
participants are measured on the dependent variable after being in each condition of the experiment.
(have the same individuals participate in all conditions)
order effect
the order of presenting the treatments affects the dependent variable. (decrease this by increasing time between conditions and with counterbalancing)
practice effect
when performance improves because of repeated practice with a task.
fatigue effect
when performance worsens as participants become tired, bored, or distracted.
contrast effect
when the response to the second condition in the experiment is altered because the two conditions are contrasted to one another
complete counterbalancing
all possible orders of presentation are included in the experiment. (factorial of the number of conditions gives you the number of orders)
partial counterbalancing
Latin square - a limited set of orders constructed to ensure that each condition appears at each ordinal position and each condition precedes and follows each condition once.
- N for even N and 2N for odd N
matched pairs design
Instead of randomly assigning participants to groups, the goal is to first match people on a crucial participant characteristic. (One twin from each pair would be randomly assigned to each condition)
straightforward manipulations
operationally define independent variables using instructions and stimulus presentations. ( manipulate the variable by presenting material to the participants)
staged manipulations
elaborate situations involving actors; at other times, they simply take the form of a cover story. Often use a CONFEDERATE
Manipulation strength
making the levels of the independent variable maximally different, while keeping everything else between the two groups the same
manipulation check
to directly measure whether the independent variable manipulation induced the intended psychological state among participants. (can provide evidence for the construct validity of the manipulation)
self-report measure
can be used to measure explicit attitudes, liking for someone, judgments about someone’s personality characteristics, intended behaviours, emotional states, attributions about why some- one performed well or poorly on a task, confidence in one’s judgments, and many other aspects of human thought and behaviour.
Behavioural measure
direct observation of behaviours. (Often recording whether a behaviour occurs or how many times it occurs)
Physiological measure
a recording of a response in the body
Sensitivity
the dependent variable should be sensitive enough to detect resulting differences between groups
ceiling effect
the independent variable might appear to have no effect on the dependent measure only bc the participants quickly reach the maximum performance level (task too easy)
Floor effect
when a task is so difficult that hardly anyone can perform well
Demand characteristics
any feature of a study that might inform participants of its purpose and consequently affect their behaviour.
filler items
to decrease demand characteristics - disguise the dependent measure by using an unobtrusive measure or by placing the measure among a set of unrelated filler items on a questionnaire.
Placebo group
participants in the placebo group receive a pill or injection containing an inert, harmless substance; they do not receive the drug given to members of the experimental group
Experimenter bias
experimenters are aware of the purpose of the study and may develop expectations about how participants should respond - these expectations can bias the results
single-blind procedure
participants are unaware of which condition they are in (e.g., whether a placebo or the actual drug is being administered)
double-blind procedure
neither participant nor experimenter knows the participant’s condition.
pilot study
when the researcher does a “trial run” with a small number of participants drawn from the same population as the sample he or she ultimately hopes to test.
levels
In the simplest experimental design, the independent variable has only two
factorial design
has more than one independent variable (factor), and all levels of each independent variable (IV) are combined with all levels of the other independent variables. (Number of levels of first IV x Number of levels of second IV)
2x2 design
Two independent variables with 2 levels and therefore 4 conditions
main effect
the effect that each independent variable by itself has on the dependent variable
Interaction
If there is an interaction between two independent variables, the way that one independent variable affects the dependent variable depends on the particular level of the other variable.
marginal mean
Tells you if an independent variable has a main effect
moderator variable
influences the relationship between two other variables
Analyzing factorial graphs
Interaction:
- Graph: If the lines are NOT parallel - there is an interaction
- Table: look at the change in values from B1A1 to B1A2 and from B2A1 to B2A2 - if different, there is an interaction
Main effect:
- Table: If the marginal means (for a single IV) are different, there is a main effect.
- Graph:
- To determine if B has a main effect, find the average of the two points on a single line to make a horizontal line - do this for both lines and if they are different, it has a main effect
- To determine if A has a main effect (x-axis), find the average of the two lines and if it has a non-zero slope, A has a main effect
simple main effect
examines the mean differences at each level of one independent variable
IV X PV design
- a factorial design that includes both experimental (manipulated) and non-experimental (measured/non-manipulated) variables
- independent variable by participant variable
- investigate how different types of people respond to same manipulated variable.
- participant variables are often personal attributes such as sex, age, ethnic group, personality characteristics, or clinical diagnostic category.
- Participant variables cannot be randomly assigned or controlled: Participants bring those characteristics with them to the study.
Independent groups
- different people assigned to each of the 4 conditions in a 2x2 design
- 2x2 and if you want 10 participants in each condition you need 40 people
Repeated measures
- the same people participate in all conditions
- 10 people needed (if 10 per condition)
Mixed factorial design
- a combination of independent groups and repeated measures
- One group for A1 that receives both levels of B
- One group for A2 that receives both levels of B
- (20 people needed if 10 per condition)
frequency distribution
indicates the number of participants who receive or select each possible score on a variable, and can be created for variables using any scale.
outliers
scores that are unusual, unexpected, or very different from the scores of other participants
bar graph
- uses a separate and distinct bar for each piece of information.
- Best for comparing group means but also for percentages
pie chart
- divides a whole circle, or «pie,” into «slices” that rep- resent relative percentages.
- Nominal scale info
Histogram
uses bars to display a frequency distribution for a continuous variable.
Frequency polygons
- an alternative to histograms, use a line to represent frequencies when the variable uses an interval or ratio scale.
- when you wanna examine frequency for multiple groups simultaneously
descriptive statistics
to make precise statements that summarize the data.
central tendency
tells us what the sample is like as a whole, or on the average.
mean
obtained by adding all the scores and dividing by the number of scores.
median
the score that divides the group in half (50% above and 50% below the median)
Mode
the most frequent score
Variability
- how widely the distribution of scores is spread
- a number that characterizes the amount of spread in a distribution of scores that are measured on interval or ratio scales
Standard deviation
indicates how far away scores tend to be from the mean
Variance
standard deviation squared
range
the difference between the highest score and the lowest score.
When is a hypothesis falsifiable?
- It takes risks
- If the test can refute it
When is a hypothesis unfalsifiable?
- When no empirical evidence is obtainable
- When its predictions are so vague that they can hardly fail; they become irrefutable
- when it is upheld even though refuted by data, by introducing additional assumptions and interpretations post-hoc
systematic error
the difference between the measure and the “true” value of that variable
accuracy
inversely related to the degree of bias
What makes a good operational definition?
- Reliability
- Absence of bias
- Cost
- Practicality
- Objectivity
- High acceptance
- Validity `
Quantitative variable
- measure a type of magnitude
- have # values
- can be distinguished with a subtraction test
Discreet vs continuous quantitative variables
- use a midway test to determine
- take 2 whole # levels and check in between them; if the midway has meaning then it is continuous and if not it’s discreet
- discreet = # of siblings
- continuous = speed of a car
non-monotonic relationship
- when the graph has a + relationship at one point and a - relationship at one point
monotonic relationship
when the graph has the same relationship the entire time and doesn’t switch direction (only + or only -)
What does correlation imply?
- the correlation is spurious (no reason for them to match)
- A causes B
- B causes A
- a third variable is causes both A and B
inferential stats
helps us generalize from the sample to the population
symmetric distribution
can be divided into two halves that are mirror images of each other
negatively skewed distribution
has score values with low frequencies that trail off towards negative numbers
positively skewed distribution
has score values with low frequencies that trail off towards positive numbers
bimodal distribution
has two peaks
unimodal distribution
has one peak
uniform distribution
does not have a well defined mode (no peak)
the law of large numbers
as sample size increases, sample stats become less variable and more closely estimate population values
Central tendencies suitable for certain data:
nominal: mode
ordinal: median, mode
interval: mean, median, mode
ratio: mean, median, mode