Field Methods

frequency

• Describe a particular rate or degree of a single variable

• Prioritizes construct validity, statistical validity, and external validity

surveys

Conducted to describe how members of a population distribute themselves on one or more variables

consensus

gathers information from every member of a population rather than a sample

cross-sectional

type of survey where data is collected from a predetermined population at one point in time

longitudinal

type of survey where data is at several points in time to track changes over time

broadest, general, objectives

Step 1: Defining the problem

• Begin with the _______, most ________ question and end with the most specific

• Clearly define ________ and each question should relate to one or more of the survey’s

data, target population

Step 2: Identifying the target population

• _____ collected will be summarized and used to describe the population this sample represents

• __________ must be clearly defined such that it is possible to tell whether any member of the sample is definitively a member of the population

inclusion

Participants must have a set of characteristics to be eligible for a study. These criteria ensure the selected sample is relevant to the research question.

exclusion

A set of characteristics that disqualify participants from a study. These criteria help eliminate confounding variables that could affect the results.

data collection

Step 3: Choosing the mode of _________

• mail surveys, telephone surveys, personal interviews, or electronic surveys

randomly

Step 4: Selecting the sample

• Selected ______ (if possible)

instrument, questionnaire, interview

Step 5: Preparing the _________

• The most common types: _________ (self-administered) & the __________

self-report

A type of research method wherein participants report their own answers to questions

open-ended

Questions that allow participants to respond in their own words.

dichotomous

questions that are Yes-No, True-False

forced choice

A question format where respondents must choose between a set of predefined options.

likert scale

A rating scale used to measure attitudes or opinions, typically ranging from Strongly Disagree to Strongly Agree.

semantic differential

A scale measuring the meaning participants associate with a concept, often using opposing adjectives.

double-barreled questions

It asks about two or more issues at the same time, making it difficult for respondents to provide a clear answer.

acquiescence bias

Tendency to agree with statements regardless of content.

fence-sitting

Choosing neutral or middle responses instead of expressing a clear opinion.

faking good/bad

Respondents give socially desirable answers to appear better or respondents exaggerate problems or negative traits.

electroencephalogram

Measures electrical activities of the brain

magnetic resonance imaging

Produces a black-and-white image of the body by using radio waves

positron emission tomography

Uses radioactive tracers which are swallowed, injected, or inhaled to reveal the functionality of a person’s tissues and organs

observational research

Systematically observing and recording behavior to describe a variable

naturalistic observation

Observing people’s public behavior in the environment in which it typically occurs

participant observation

Researchers observe behavior by being active participants in the group or situation they are studying

structured observation

Observing behaviors either in a laboratory or using some degree of manipulation

archival research

• Considered an observational method; analyzes existing data that has been collected for some other purpose

• World Health Statistics, Open Data Philippines, World Bank Open Data, Dataset Research

construct validity

Ensures the measured variable accurately represents the concept.

rosenthal effect

researcher's expectations subtly influence participants' responses or behavior.

Ex: If a teacher expects certain students to perform better, they may unconsciously give them more encouragement, leading to higher performance.

Threat to C.V.: The results may reflect the experimenter's expectations rather than the actual abilities or traits being measured

experimenter effect

• Any unintended influence the researcher has on the study.

• due to tone of voice, body language, wording of questions, or even the presence of the researcher affecting participants' behavior.

• Ex: A researcher conducting a survey on anxiety might unintentionally show concern when asking certain questions, making participants more likely to report higher anxiety.

• Threat to C.V.: The results may reflect the researcher's influence rather than the actual construct being studied.

pygmalion effect

high expectations lead to improved performance.

Ex: If a manager believes an employee is highly competent, they may give them better tasks and feedback, leading the employee to perform better.

Threat to C.V.: The improvement may not be due to actual ability but rather the positive reinforcement received, making it unclear whether the measured trait (e.g., intelligence, skill) is real or externally influenced.

single-blind study

- Participants do not know which experimental condition they are in.

- This prevents participant bias, such as the placebo effect, where participants change their behavior based on expectations rather than actual treatment.

- Ex: In a drug trial, participants don’t know whether they are receiving the real drug or a placebo.

double-blind bias

• Neither the participants nor the experimenters know who is in the experimental or control group.

• This eliminates both participant and experimenter bias

• Ex: In a medical study, neither the doctor nor the patient knows whether they are receiving the real medication or a placebo.

statistical validity

Ensures the findings are statistically sound and reliable

internal validity

(Less relevant for frequency claims), as it focuses more on causality.

external validity

Determines whether the results can be generalized to other populations or settings.

observer bias

observers see what they expect to see

observer effects

Participants confirm observers’ expectations.

reactivity

People change their behaviors when they’re being watched.

bivariate correlation

- investigating the association between two variables

- Can be used to explain and predict relationships

causal-comparative

- also called ex post facto research

- Investigates difference between two groups without manipulation

- No randomization & controlled design

surveys, neuroimaging, archival

types of data gathering a causal-comparative can apply

type 1 error

- False positive

- Rejecting the null hypothesis when it is actually true.

- “Congrats you’re pregnant.” to a man

type 2 error

False negative

- Failing to reject the null hypothesis when it is actually false.

- “You’re not pregnant.” to a pregnant woman

effect size

- Magnitude/strength of relationship

- Larger are usually important; allows more accurate predictions

5

if outliers constitute less than _% of the data, their overall impact might be minimal.

range restriction

When _________ occurs, researchers can correct it statistically or recruit more people at the extremes.

Example:

Interpretation:

Until Grade 3 = no association/pattern

Until Grade 12 = wide range (can be seen in the pattern)

curvilinear relationship

- As one variable increases, the other may rise then fall, or vice versa.

- SPSS cannot detect this.

covariance, temporal precedence, internal validity

three criteria for causality

covariance

Results must show a correlation between cause variable and effect variable

temporal precedence

The cause variable must precede the effect variable in time.

internal validity

No plausible alternative explanations should exist for the relationship between two variables.

multiple regression

computes the relationship between a predictor and a criterion variable, controlling for other predictor variables; also called multivariate regression

computer simulation

also called modeling; a method which implements a theoretical model in a computer

trend, cohort, panel

kinds of longitudinal studies

trend studies

longitudinal survey where population changes overtime

cohort studies

longitudinal survey where it is the same population with different sample

panel studies

longitudinal survey where it is the same sample from the same population throughout time.

cross-sectional correlational

• examines the relationship between two or more variables that are measured at the same time (i.e., in a single snapshot).

• It’s a type of non-experimental research that looks at associations, not causation.

cross-lag correlations

• analyze how one variable at an earlier time point predicts another variable at a later time, while also considering their earlier correlations.

• This is done using at least two time points and two variables.

criterion, predictor

when interpreting regression results, ________ is the outcome you're trying to predict (e.g., stress level) while ________ is the factor you think influences the outcome (e.g., hours of sleep).

beta

- This tells you how much the criterion changes when the predictor changes by 1 unit.

- It can be positive (increases outcome) or negative (decreases outcome).

absolute value

when interpreting the beta, a larger ________ means a stronger effect

positive, negative

- _______ β: As the predictor increases, so does the criterion.

- _______ β: As the predictor increases, the criterion decreases.

unstandardized coefficient

Shows how much P_TOT changes when each predictor goes up by one unit.

standardized coefficient

Helps you compare which predictor has the strongest effect, even if they’re on different scales

p-value

The t-value is a test statistic used to calculate the ________. A low t-value (close to 0) means the predictor is not strongly contributing to the outcome.

mediating variable

A _________ helps explain the relationship between two other variables and they ideally establish temporal precedence

moderating variable

a variable that, depending on its level, changes the relationship between two other variables