ARM Exam 1 Cards with NO HINTS

Research

A systematic investigation designed to develop or contribute to knowledge that can apply beyond the specific cases studied.

Overgeneralization

Concluding that something seen in only a few cases applies to an entire group.

Ex: Meet two rude UM students → conclude all UM students are rude.

Incorrect Observation

Perceiving or recording something inaccurately and then drawing a conclusion from that mistake.

Ex: A researcher counts 20 people at a protest when there were actually 30, then uses the incorrect count in the study.

Selective Observation

Paying attention only to evidence that supports an existing belief while ignoring evidence that contradicts it.

Ex: Believe immigrants increase crime → notice only stories about immigrant crime.

Illogical Reasoning

Reaching a conclusion on the basis of assumptions that do not validly support it.

Ex: A student studies for an exam while wearing a red shirt and gets an A, so they conclude the red shirt caused the good grade.

Resistance to Change

Continuing to hold existing ideas even after being presented with new information that challenges them.

Ex: Someone believes crime is increasing and continues believing it even after being shown reliable data that crime has decreased.

Descriptive Research

Examines a phenomenon primarily by determining how much, how many, or what characteristics are present.

Ex: Ask, “How many college students use Instagram?”

Exploratory Research

Investigates what is happening when there are not yet strong expectations about what will be found.

Ex: Interview students to learn why they are leaving college.

Explanatory Research

Investigates why something occurs by identifying causes and effects.

Ex: A researcher studies whether education level affects income to explain why some people earn more than others.

Evaluation Research

Determines the effects or outcomes of a social program or policy.

Ex: Study whether a new after-school program reduces delinquency.

Quantitative Research

Uses numerical data and statistical analysis to investigate questions.

Ex: Survey 500 people and statistically analyze their responses.

Qualitative Research

Uses primarily text-based or nonnumerical information to develop an in-depth understanding.

Ex: Interview 20 people about their experiences.

Mixed Methods

Combines numerical and text-based approaches within the same investigation.

Ex: Interview some people and survey others (lecture notes example).

Research Question

A specific question about a topic that an investigator seeks to answer.

Confidential Data

Personal identifying information is collected but is protected from being shared.

Anonymous Data

Personal identifying information is never collected.

Voluntary Participation

Individuals take part by their own choice and without coercion.

Informed Consent

Agreement to take part after receiving information about the study and its possible risks and benefits.

Assent

Agreement to take part obtained from a child who is not legally able to provide full adult consent.

Deception

Withholding or misrepresenting certain information when necessary for a minimal-risk study.

Debriefing

Providing participants with previously withheld or misrepresented information as soon as possible after their participation.

Institutional Review Board (IRB)

A body responsible for reviewing studies involving human participants and granting approval before data collection begins.

Triangulation

Studying the same topic through multiple sources, methods, or investigators to increase confidence in the findings.

Ex: Ask several people the same questions and compare their accounts (lecture notes example).

Ontological Assumption

A philosophical position concerning what is real and the nature of reality.

Epistemological Assumption

A philosophical position concerning the relationship between the investigator and what or whom is being studied.

Axiological Assumption

A philosophical position concerning the role that values play in the investigation.

Methodological Assumption

A philosophical position concerning how the process of investigation should be conducted.

Deduction

Beginning with a broad theory and moving toward specific hypotheses and observations.

Think: Theory → hypothesis → collect data.

Induction

Beginning with specific observations and patterns and moving toward broader theoretical explanations.

Think: Observe a pattern in data → develop a theory.

Unit of Observation

The level or entity at which information is collected.

Ex: Study jails; an employee answers questions about the jail, but the jail is the unit (lecture notes example).

Informant

A person who supplies information about the entity being studied.

Ex: A jail employee provides information about the jail (lecture notes example).

Attribute

A characteristic or quality belonging to a single case.

Ex: Male is an attribute of the variable sex (lecture notes example).

Variable

A logically grouped characteristic whose value can differ across members of a population.

Ex: Gender varies across people (lecture notes example).

Constant

A characteristic that has the same value for every member of the population being studied.

Ex: At an all-male college, sex is constant (lecture notes example).

Level of Measurement

The amount and type of information contained in the values assigned to a characteristic.

Nominal

Values place cases into distinct categories but provide no meaningful ranking among those categories.

Ex: Race: White, Black, Other.

Ordinal

Values place cases into categories that can be meaningfully ranked, although the distances between ranks are unknown.

Ex: Satisfaction: dissatisfied → neutral → satisfied → very satisfied.

Interval

Values can be ranked and have known equal distances between them, but zero does not represent the complete absence of the characteristic.

Ex: Celsius temperature: 0°C does not mean no temperature.

Ratio

Values can be ranked, have known equal distances between them, and include a zero that represents the absence of the characteristic.

Ex: Number of children: 0 means no children (lecture notes example).

Mutually Exclusive

The categories are arranged so that any single observation can belong to only one of them.

Ex: Age groups 18–29, 30–39, 40+: a 35-year-old fits only one.

Mutually Exhaustive

The categories are arranged so that every possible observation has somewhere to be classified.

Ex: Age groups cover every person in the sample.

Dichotomous Variable

A characteristic that can take exactly two categories or values.

Ex: Yes/No.

Polytomous Variable

A characteristic that can take three or more categories or values.

Ex: Race: White, Black, Asian, Other — the variable has more than two categories.

Dependent Variable (DV)

The outcome that an investigator is attempting to explain.

Ex: Studying whether education affects income → income is the DV because it is the outcome being explained.

Independent Variable (IV)

The predictor or explanatory factor thought to influence an outcome.

Ex: Studying whether education affects income → education is the IV because it is used to explain income.

Positive Relationship

Higher values of one characteristic are associated with higher values of another, or lower values are associated with lower values.

Ex: More education → more income.

Negative Relationship

Higher values of one characteristic are associated with lower values of another, and vice versa.

Ex: More education → less unemployment.

Bivariate Relationship

An association involving exactly two characteristics.

Ex: Study the relationship between education and income. There are exactly two variables: education and income.

Multivariate Relationship

An association or analysis involving three or more characteristics.

Ex: Study how education, work experience, and gender relate to income. The relationship involves four variables.

Direct Relationship

One factor affects an outcome without operating through an intervening factor.

Ex: Study hours → exam score.

Indirect Relationship

One factor affects an outcome by first influencing another factor that then influences the outcome.

Ex: Education → occupation → income. More education can lead to a higher-paying occupation, which then leads to higher income. Education affects income through occupation.

Recursive Relationship

Causal influence proceeds in only one direction, such as X → Y.

Ex: Years of education → starting salary. Education is completed first and can affect later starting salary; the later salary cannot go backward and change the education already completed.

Non-recursive Relationship

Two factors can influence one another, creating a feedback loop such as X ↔ Y.

Ex: Stress ↔ poor sleep.

Total Effect

All pathways by which X influences Y considered together.

Ex: Education affects income directly and through occupation; combine both effects.

Deterministic Causality

The presence of X always produces a particular Y outcome, or never produces it.

Ex: Being beheaded → death. The cause produces the outcome, not merely increasing its probability.

Probabilistic Causality

The presence of X changes the likelihood of Y occurring without guaranteeing that Y will occur.

Ex: Males are more likely to commit crime, but not all males commit crime (lecture notes example).

Correlation

An empirical association must exist between X and Y.

Ex: As education increases, income also tends to increase.

Temporal Order

A proposed cause must occur before its proposed effect.

Ex: Sex exists before later criminal behavior (lecture notes example).

Non-spuriousness

An association between X and Y must remain after prior third-variable explanations are considered.

Ex: Suppose education is related to income. After controlling for work experience, education is still related to income. Work experience does not explain away the education–income relationship.

Parsimony

Explaining a phenomenon using the most important factors without adding unnecessary complexity.

Ex: To explain income, a researcher uses education and work experience because theory says they matter, rather than adding dozens of unrelated variables such as favorite color or favorite food.

Measure of Association

A numerical indication of the magnitude or strength of the connection between characteristics.

Ex: A strong relationship between education and income.

Statistical Significance

The probability that an observed association could be attributed to chance.

Ex: A study finds that education is related to income with p < .05, meaning the observed relationship would be unlikely under the null hypothesis.

Exogenous Variable

A factor in a model that is not caused by another factor included in that model.

Ex: Age → attitudes; age is exogenous if nothing in the model causes age.

Endogenous Variable

A factor in a model that is caused or affected by another factor included in that model.

Ex: Education → income; income is endogenous.

Spurious Relationship

An apparent connection between X and Y that exists because an earlier third factor influences both.

Ex: Hot weather → ice cream sales AND drownings; ice cream does not cause drowning.

Specification Error

A modeling mistake caused by choosing the wrong factors or incorrectly representing how they are connected.

Ex: Predict Trump support with age, education, and income but leave out political ideology (lecture notes example).

Type 1 Specification Error

A modeling mistake caused by leaving out an important factor that should have been included.

Ex: Leave political ideology out when it belongs in the model.

Type 2 Specification Error

A modeling mistake caused by including an irrelevant factor that should have been left out.

Ex: Put PB&J preference in a model when it is irrelevant (lecture notes example).

Type 3 Specification Error

A modeling mistake caused by representing the form of a connection between factors incorrectly.

Ex: The true relationship between age and crime is curved—crime rises and then falls with age—but the researcher models it as a straight-line relationship. The variables are right, but the form of the relationship is wrong.

Conceptualization

Establishing precisely what an abstract idea means for purposes of a study.

Ex: Stress = feeling overwhelmed or unable to cope.

Operationalization

Deciding the specific procedure or indicator that will be used to turn an abstract idea into something observable.

Ex: Measure stress by asking, “How often did you feel overwhelmed this month?”

Measurement

Carrying out the data-collection procedure after deciding how an idea will be represented empirically.

Ex: Actually collect respondents’ answers to stress questions.

Proposition

A theoretical statement describing an expected causal connection between abstract concepts.

Ex: Higher SES → lower religiosity (lecture example).

Hypothesis

A testable statement describing an expected causal connection between measurable variables.

Ex: Higher income → lower church attendance (lecture example).

Interchangeability of Indicators

Different empirical indicators representing the same underlying concepts should generally produce similar patterns of relationships.

Ex: Measure SES with income, education, or occupation and get similar patterns (lecture example).

Dimensionality

The extent to which an abstract concept consists of a single component or several distinct components.

Ex: Teacher stress can include workload, finances, health, and sleep.

Reliability

The extent to which repeated applications of a measure produce stable and consistent results.

Ex: Scale says 150 every day when you actually weigh 160: consistent but wrong.

Cronbach's Alpha

A 0-to-1 coefficient based on relationships among multiple items, with higher values indicating that the items function more consistently together.

Validity

The extent to which an instrument actually captures the concept it is intended to capture.

Ex: A valid scale measures your actual weight accurately.

Face Validity

Whether an instrument appears, based on its content, to capture the concept it is intended to represent.

Ex: Depression survey asks about sadness and hopelessness.

External Validity

Whether findings obtained from the studied cases can be generalized beyond those cases to a broader group.

Ex: Results from a representative U.S. sample generalize to U.S. adults.

Criterion-Related Validity

Whether results from one instrument agree with results from another relevant instrument used as a comparison standard.

Ex: New depression test gives similar results to an established test.

Concurrent Validity

Whether one instrument agrees with another relevant instrument when both are administered at approximately the same time.

Ex: Give someone a new depression test and an established depression test on the same day. If the scores are similar, the new test has concurrent validity.

Predictive Validity

Whether scores obtained at one point can accurately anticipate a later outcome.

Ex: Entrance-exam scores predict later college performance.

Content Validity

Whether an instrument adequately covers all important components of the concept it is intended to capture.

Ex: A test of math ability should include different areas of math, such as algebra and geometry. If it asks only algebra questions, it does not cover the full concept of math ability.

Construct Validity

Whether an instrument behaves and relates to other factors in the way that theory says it should.

Ex: If a scale really measures depression, people scoring higher on it should also tend to report more depressive symptoms, as theory predicts.

Test-Retest Reliability

Whether the same instrument administered to the same cases at different times produces similar results.

Ex: Give the same questionnaire today and again two weeks later; answers are similar (lecture example).

Internal Consistency

Whether multiple questions intended to capture the same underlying concept produce compatible results.

Ex: Someone high on one stress question is also high on the other stress questions.

Null Hypothesis

A statistical statement proposing that no association exists between the variables being examined.

Alternative Hypothesis

A statistical statement proposing that an association exists between the variables being examined.

Type I Error (Alpha)

Concluding that an association exists when in reality it does not; a false positive.

Ex: Conclude education and income are related when they actually are not. False positive.

Type II Error (Beta)

Failing to detect an association that actually exists; a false negative.

Ex: Conclude there is no education-income relationship when one actually exists. False negative.

Pearson's Chi-Square Test of Independence (χ²)

A statistical procedure used to determine whether two categorical variables are associated rather than unrelated.

Chi-Square Formula

χ² = Σ (O − E)² / E, where O represents the value found in the data and E represents the value anticipated under independence.

Expected Value Formula

(Row marginal × column marginal) / total.

Chi-Square Critical Value (1 df)

With one degree of freedom, 3.84 is the cutoff at the .05 level; values at or above it meet the criterion while values below it do not.

Degrees of Freedom (df)

A quantity used with a test statistic to determine the threshold it must reach for the result to meet the chosen criterion.

Inferential Statistics

Procedures that use information from a subset of cases to estimate, predict, or draw conclusions about a larger group.

Ex: Use a sample of 500 voters to make an inference about all voters.

Descriptive Statistics

Procedures used to organize, depict, and summarize the characteristics of collected data; mean, median, and mode are examples.

Data Matrix

An arrangement of a dataset in which rows correspond to individual cases and columns correspond to characteristics.