Research Methods Exam 1

empirically

conclusions based on systematic observations (scientific method)

empiricism

evidence derived from senses

producer

conduct research to answer a scientific question

produce information

consumer

reads, interprets, applies scientific findings

ex: teacher, intervention evaluator

Merton’s Scientific Norms

1) empiricism—>universalism, disinterestedness, organized skepticism

2) examines solvable problems

3) producing public knowledge

universalism

scientific claims are evaluated according to their merit

communality

scientific knowledge is created by a community and findings are for community

disinterestedness

scientists strive to discover truth; not swayed by anything

-not personally invested in income

organized skepticism

scientists question everything, including their own theories

theory data cycle

theory

set of statements that describe general principles about how variables relate to each other

hypothesis

-a falsifible prediction

-specific outcome that researchers expects to observe in a study if theory is accurate

data

set of observations

pre-registered

before collecting any data, researcher publicly states what the study outcome is expected to be

how do scientists approach work

1) single theory—>endless hypothesis

2) single study does NOT test an entire theory, only part

3) researchers test theories with senses of studies, each addressing different hypothesis

4) when data does NOT match—> theory needs to be revised, or design should be improved

good theories

• supported by data

• theory supported by LARGE QUANTITY and VARIETY OF EVIDENCE

• are testable

• have parsimony (simplest answer is the best)

Do theories prove anything?

NO, “supported” or “not supported”

What are the two types of research?

1) applied research

2) basic research

applied research

address a particular problem, applying a method and seeing how it helps

basic research

enhance the general body of knowledge, engine of evidence

translational research

bridge between basic and applied research

example: basic research on cell membrane biochemistry might be translated into a new drug for mental health issues

Intro for APA

cites past research to explain rationale for current hypothesis

Methods for APA

• recipe for study

• participants, design, materials, procedure

Results APA

data analyses

Discussion APA

theoretical implications, limitations, future directions

steps for the publication process

1) scientists share research by submitting a paper

2) peer review stage: send paper to ¾ experts in the field to review

3) the process is rigorous and anonymous. manuscripts are either rejected, accepted, or accepted with revisions

two important safeguards for ethics

1) ethics codes

2) institutional review boards

Belmont Principles

1) respect for persons

2) beneficience

3)justice

respect for persons

• must be treated as autonomous individuals

1) informed consent

2) make research voluntary

3) use special protections for participants with limited autonomy

beneficence

• maximize benefits of research while minimizing harm to participants

1) use research methods with least possible risk

2) maintain participant confidentiality

3) carefully monitor research that involves more than minimal risk

justice

• procedures need to be administered fairly among participants

1) treat research participants equitably

2) do NOT exploit vulnerable populations

passive deception (deception of omission)

withholding of truth/relevant info

active deception (deception of commision)

intentionally misinforming

Institutionalized Review Boards (IRBS)

• board reviews research proposals

• must have at least 5 members

• 1 member must be outside of institution

what is a variable?

something we manipulate or measure, something that varies

“More than 2 million U.S Youths are depressed”

variable: depression

levels: depressed and not depressed

measured variable

• one that is observed and recorded

• some are measured using familiar tools (IQ test, ruler)

• others need a measurement to be created (gender, hair color)

manipulated variables

• one that researcher controls

• assigning participants to the levels of that variables

random assingment

random placement of ppts into levels of manipulated variable

constant

something that has potential to vary, but does not in a given study

example: in research on fathers, sex is constant rather than a variable

conceptual definition

a researchers definition of a variable at an abstract level (the WHAT)

operational definition

the specific way in which a concept is measured or manipulated (the WHY)

what are the 3 types of claims?

1) frequency

2) association

3) casual

frequency claims

• rate/degree of a particular thing or event

• how frequent or common something is

• usually a percentage

• example: 1=never, 2=alittle and so on

association claims

• displayed on scatter plots

• on variable plotted on x-axis, one variable plotted on y-axis

• 4 types of associations:

  • Positive

  • As X increases, Y increases

  • Negative

  • As X increases, Y decreases

  • Zero

  • Knowing X tells us nothing about Y

  • Curvilinear

  • As X increases, Y changes its pattern

  • E.g. Y increases, then decreases

Casual Claims

• argues that one variable is responsible for changing another

• changes in “x” CAUSE changes in “y”

• positive, negative, curvilinear

correlation does NOT cause

CAUSATION

third variable problem

there may be an unmeasured variable that actually causes variables to covary (change together)

directionality problem

not always possible to specify the direction in which a casual arrow points

do video games and internet ruin our relationships?

What are the four types of validities?

1) construct

2) statistical

3) external

4) internal

construct validity

how well did a study measure or manipulate a variable?

ex: 39% of teens text while driving

statistical validity

• researchers use statistics to analyze their data

• are the results statistically accurate?

• how well do the numbers support the claim?

external validity

• how well do the results generalize?

• can our findings apply to other populations, contexts, times or places?

internal validity

how likely is it that one of the variables cause the change in the other?

what are the 3 criteria to establish criteria claims…?

1) covariance

2) internal validity

3) temporal precedence

covariance

are the variables associated with each other?

internal

was the study an experiment?

temporal precedence

did one happen before the other?

random selection

whole population of interest has an equal chance to be in study

random assignment

participants are randomly assigned to a level

categorial measurement

• levels are categories

• ex: spring, summer, fall, winter

• ex: junior, senior, freshmen

quantitative measurement

• continuous

• meaningful numbers

• inherently numerical

ordinal scale (quantitative variable)

• ranked scale

• uneven intervals

• ex: top 10 selling books

ratio scale (quantitative variables)

• equal intervals

• “true 0” (0 means none)

• ex: number of exam questions correct

interval scale

• equal intervals

• no “true 0” (0 doesnt mean none)

• ex: IQ test

what are the common types of measures?

1) self-report measure

• questionarre/interview

2) observational measure

• observable behavior or traces of observable behavior

3) physiological measure

• biological data (heart rate)

test-retest validity

• scores should be consistent each time they are measured

• test at a specific time, then test a few years later

• most relevant for stable constructs

• ex: IQ test

Interrater reliability

• when we have multiple observers

• consistent scores despite different observers

• ex: 2 raters count how often a child smiles on a playground

Internal reliability

• consistent pattern of answers despite phrasing of questions

• applies to measures that combine multiple items

face validity

• it looks like something you want to measure

• does it pass the “vibe” check

content validity

• the measure contains all the parts that your theory says it should contain

criterion validity

• the measure is associated with a concrete behavioral outcome it is expected to be

• look at correlational evidence

• known-groups paradigm

convergent validity

does the self report measure correlate with other measures of the same construct

discriminant validity

does the self report measure NOT correlate with measures of dissimilar constructs