Research Methods exam vocab

subject variable

A variable that is inherent to the participants whose effect is measured rather than manipulated, e.g. gender

Supersitition

Way of knowing: beliefs contrary to reason

Intuition

Way of knowing: due to gut feeling rather than conscious reasoning, but not irrational

Authority

Way of knowing: trusting a credible source

Rational-inductive argument

Type of knowing: an argument or opinion formed based on facts, previous experiences, context, etc. Used in humanities.

Scientific method

Way of knowing: “top-down,” for empirical questions that can be answered with observation. Ask a q, background research, hypothesis, test, write it up, replicate or restructure

Basic research

Conceptual-based research that seeks to establish general trends or truths without immediate practical applications

Applied research

Research with a specific and immediate real-world application

3 characteristics of science

1. Systematic empiricism - what it sounds like

2. Publicly verifiable knowledge - open to refution, replication, peer-review

3. Addresses solvable problems - can be proven wrong

True experiment

1. Researcher manipulates at least one IV

2. Participants are randomly assigned to conditions

Belmont Report

A report establishing the three principles of ethical research: beneficence (minimize harm, maximize benefits), respect for persons (informed and voluntary consent), and justice (risks and benefits should be fairly distributed)

Ethical principles for working with data

Honesty, carefulness, confidentiality

4 types of risk

Physical, social, psychological, economic

IRB

Institutional Review Board - an institution-specific committee of diverse experts or community members who screen and approve research proposals

Title page

Parts of a research report: where the title and authors are

Abstract

Parts of a research report: Brief summary of entire report

Introduction

Parts of a research report: Literature review and research questions

Method

Parts of a research report: how was the question explored

Discussion

Parts of a research report: what do the results mean? was the hypothesis supported? future implications?

Results

Parts of a research report: organized data from the study/what the study found

References

Parts of a research report: citations

Times New Roman 12

APA font

Hypothesis

A claim about how you expect your constructs to be related

Frequency claim

Something that describes the level of a single variable - how often something happens or the proportion of a condition in a population

Association claim

A claim establishing a definable relationship between variables - covary/correlation

Causal claims

Claims that changes in one variable cause measurable changes in another variable

Conceptual definition

A researcher’s abstract definition of a concept

Operational definition

A researcher’s specific, measurable definition of a concept within the framework of a specific study

risk

probability of harm

minimal risk

when a study has no greater risk of harm than posed by someone’s everyday life

levels of risk

minimal, minor increase over minimal, moderate, high

Objective/physiological measures

Quantifiable or observable physiological responses like heart rate, brain activity, body temp., etc.

Behavioral measures

Observable participant behavior with standardized/agreed upon definitions for researchers

Self-report

when participants self-report their experiences

reliability of measurement

how consistent a measure is

validity of measurement

how well something measures what it is supposed to be measuring

face validity

how conceivably something is measuring what it is supposed to be measuring

Scales of measurement: nominal

Numbers have no meaning, data is strictly categorical, includes yes/no

Scales of measurement: ordinal

Data categories can be placed in ascending order, but are not a standardized distance apart and do not have a true 0 (ratios cannot be made) (e.g. sizes of cups)

Scales of measurement: interval

Data categories can be ordered, are a set length apart, but have no true 0 (e.g. self-esteem scale)

Scales of measurement: ratio

Data categories are ordered, have a set width, and have a true 0 (e.g. ages, weights)

Construct validity

Measurement validity

How to assess construct/measurement validity

head, peer review, expert judgment, stats

test-retest reliability

way to measure measurement reliability - testing same participants at different points to see how similar answers are

Internal consistency/reliability

Consistency of results across various items measuring the same concept in a study - inter-item reliability

Convergent validity

How closely a measure correlates with related concepts that it should covary with

Discriminant validity

The extent to which a measure correlates with unrelated concepts/measures - if it correlates strongly, it lacks validity

Dependent measure sensitivity

The degree to which a measurement tool for the DV can detect subtle changes in groups caused by the I.V.

Ceiling effect

When all the participant data is clustered near the top of the measure, making it impossible to see potential I.V. effects - e.g. when a test is too easy, when everyone is happy by a certain measure, etc.

Floor effect

When all participant data is clustered at bottom of a measure, measure isn’t sensitive enough to find subtle differences - e.g. when a test is too hard, if a self-esteem test had only extreme measures, etc.

Internal validity

Certainty that IV caused DV - achieve with random assignment

External validity

How generalizable a study’s results are - random selection promotes this, convenience sampling decreases it

Ecological validity

Whether a study’s findings can generalize to other settings

Population validity

Whether a study can generalize to other groups of people

Reliability (study design)

How replicable a study is

Confounds: experimenter effect

When different experimenters collect data or affect outcomes differently

Confounds: experimenter bias

When experimenter’s expectations or desires impact outcomes

Demand characteristics

Elements of the study that may hint at desired outcomes/hypotheses to participants, influences their behavior (good v.s. negative participants)

Random error

natural deviations in data, “noise,” spread across groups by random assignment

Systematic error

Confounding/nuisance variables that change DV in lockstep with IV

Systematic variance

IV