Research Methods

Modules 4 - 8 in textbook

hindsight bias

the tendency to believe, after learning an outcome, that one would have forseen it

overconfidence

we tend to think we know more than we actually do

percieving order in random events

we are prone to percieve patterns in our natural eagerness to make sense of the world, random sequences often dont look random

critical thinking

thinking that does not blindly accept arguments and conclusions. examines assumptions, assesses the souse, discerns hidden values, evaluates evidence, and assesses conclusions

theory

an explanation using an integrated sense of principles that organizes observations and makes predictions

hypothesis

a testable prediction, often implied by a theory

operational definition

a carefully worded statement of the exact procedures (operations) used in a research study, how a concept or phenomenon will be directly observed + measured in a research study

replication

repeating the essence of a research study, usually with a different set of participants in different situations to see whether the basic findings extends to other circumstances

case study

• examines one individual or group in depth

• goal is to identify universal principles or make general conclusions

• useful when phenonmen being studied is rare

• researchers must beware that what is true for one indiviudal or group may not be true for all

quantitative research

• numbers

• uses numerocal data to represent degrees of a variable

qualitative data

• letters

• in depth, narrative data gathered from individuals

experimental methodology

• uses independent and dependent variables

• random assignment

non-experimental methodologies

• case study, correlation, naturalistic observation, meta-analysis

• describes behaviors

• identify relationships between variables

• CANNOT PROVE CAUSATION

Naturalistic observation

• aka field study

• oberservation of behavior in its authentic context

• researcher does NOT ineract with subjects and only watches what occurs

• Advantages:

  • shows what happens in the real world

  • useful when behavior being studied is highly dependent on the context in which it occurs

• Major Disadvantages:

  • no experiemental control, canot identify WHY behavior occurs

  • observing behavior may change the behavior

Survey

• TOOL not a method

• used to identify attitudes or reported behaviors of a large group of people

• can obtain a large amount of info in a short amt of time

• can be swayed MANY ways

Ways surveys can be swayed

• how the items are worded

• sampling bias

• researcher biases and how those influence the interpretation of results

• self report bias

• social desiarbility bias

self report bias

discrepancies exist between what people SAY they do versus what they ACTUALLY do

social desiarbility bias

people respond based on how they think the researcher wants them to answer

meta-analysis

statistical method that combines results from multiple independent studies on the same question

correlation

a measure of the extent to which two variables change together, and thus of how well either variable predicts the other

correlation coefficent

a statistical index of the relationships between two variables

• helps us figure out how closely two things vary together

• from -1.0 to +1.0

• the closer the score gets to -1 or +1, the stronger the correlation

scatterplots

reveal correlation or the lack of correlation

illusonary correlation

• the PERCEPTION of a relationship where NONE exists

  • we believe it is there, but it is not in reality

  • helps explain superstitious beliefs

experimental group

group exposed to the indepent variable/change

control group

group NOT exposed to the independent variable, serves as a comparison for evaluating the effectiveness of the treatment

independent variable

• aka manipulated variable

• what is changed in an experiment

• applied ONLY to the experimental group

dependent variable

• may/will change based on changes to the independent

• MUST be operationally defined and QUANTITATIVE

• amount is UNKOWN at the start of the study, that is what you are trying to figure out

double blind procedure

in an experiment, where neither the researchers nor the participants know which group recieved the treatment or the placebo, commonly used in drug eval studies

random assignment

assigning participants randomly, using chance, to minimize any differences between the groups + make them as equal as possible

placebo

experimental results based on expectations alone; any effect on behavior caused by the administration of an inert substance or confition, which the recipent ASSUMES is an active agent

validity

• the key goal of experimental design

• the extent to which a test or experiment meausres or predicts what its supposed to do

correlational studies

• attempt to find a connection between two variables

• CANNOT PROVE CAUSATION

effect size

• a meausre of the stength of the relationship between two variables

  • HOW MUCH one variable affected the other in a study

  • HOW MUCH variable a effects variable b in the real world

• the larger the effect size, the more one variable can be explained by another

• gives an indication of the real world significance of a study’s results

directionality problem

the results show that two variables are connected, but cannot tell us which caused the other

third variable problem

some other variable causes BOTH variables

regression toward the mean

• extreme results are rare

• future performance trends toward average level of performance (most of the time, what it would be)

experiments

• seek to find a casual relationship between two variavles

  • when variable A happens, variable B happens

• causation can be determined because all variables are help constant/controlled - except variable of interest

• downside: high experimental control means less authentic or applicable to real world scenarios

null hypothesis

• the ASSUMPTION that no difference exists between the experimental + control group

• rejected when results are unlikely to fit

participants

• population - all those who are being studied

• representative sample - the sample reflects the demographics of the whole population

• random sampling - all members of the population have an equal chance of being slected

• convienience sampling - the sample was the most readily available

• sampling bias - the sample is not the representaative of the whole population

all experiments have:

• hypothesis

• null hypothesis

• participants

• random assignment

• control group

• experimental group

• independent variable

• dependent variable

confounding variable

• other factors that COULD influence results

• proper sampaling + blind procedurs control these as much as possible

single blind procedure

the participants do not know whether they are in the experimental group or control group, but the researchers do

measures of central tendency

a single score that represents a whole set of scores

• ex: mean - average score

  • median - middle score

  • mode - most occuring score(s)

measures of variation

shows how similar or diverse the scores actually are

• ex: range - the diff between the highest and lowest scores

  • standard deviation - how much scores vary around the mean

normal curve

• a symmetrical, bell shaped curve that describes the distribution of many types of data

• mean, median, and mode are at the same place

inferential statistics

numerical data that allow one to generalize - to infer from sample data the probability of smthn being true of a population

effect size numbers

• small effect size - 0.2 and below

• medium - 0.3 to 0.7

• large effect size - 0.8 or higher

quasi experiment

an experiment in which the participants CANNOT be randomly assigned for the purposes of comparing

• ex: age, gender, ethnicity studies

normal curve

• mean, median, and mode are ALL THE SAME

• represents a perfectly symmetrical distribution

culture

• our culture shapes our behaviors

  • collectivist - emphasizes GROUP goals

  • individualist - priority on INDIVIDUAL goals

gender

• impacts our behavior

• gender diffs in how we dream, likelihood of disorders, how we express emotions + more

statistical significance

• results are likely NOT due to chance variation

• calculation is the p-value

  • less than 0.05, it is statistically significant

  • less than 0.001, the results are statistically highly significant

ethics w/ animal research

• studies should have clear purpose

• animals must be treated humanely (APA guidelines)

• animals must be acquired LEGALLY

• least amt of suffering possible

ethics w/ human research

• informed consent

• debriefs (especially when there is deception)

• IRB (institutional review board) MUST approve before researcher begins

  • more review for studies involving children or other vulnerable pops.

• confidentiality or anonymity

• parental consent if minors

• protect subjects from harm

Anonymity

• no personal identifiers (such as names, emails, etc) are ever collected

• impossible for anyone (including researcher) to link a response to a specific person

confidentiality

• the researcher knows the participants’ identities but does NOT share them

• reasercher seperates personal data from study results and promises to never share names or identifying details