Research Methods Section 3

Experimental Design

Independent Variable

The manipulated variable in a study

Dependent Variable

The measured variable in a study, outcome.

Control Variable

Any variable that an experimentor holds constant.

Confounds

2and variables that vary systematically within the indepedent variable and obscure the result.

Requirments for causality

Temporal precedence, Covariance, Internal validity

Systematic Variability

the predictable, non-random difference in data caused by specific, identifiable factors or experimental manipulations

Unsystematic Variability

Random differences that effect both groups

Selection Effect

When participants in one level of the IV are systematically different thant those in the DV

Random Assignment

All participants have equal chance of selection

Matched Groups

matching participants based on scores, top 2 together, bottom 2 together ect. and then randomly assigning them to each group. Person A of pair 1 to group 1 and Person of pair 1 B to group 2

Between Group Design

Each participant receives 1 level of the manipulation (IV)

Within Group Designs

Each participant receives all levels of the manipulation (IV)

Between Group Posttest Design

Tested post condition is inflicted

Between Group Pretest Posttest Design

Tested both before and after condition is inflicted.

Within Group Repeated Measures Design

Partcipant receives all measures

Within Group Concurrent Measure Design

Exposed to all levels of the manipulation (IV) at the same time, testing for preference (DV). Infants

Order Effects

Being exposed to one condition affects how participants respond.

Practice Effects

Participants get better with practice.

Fatigue Effects

Particiapants get worse due to fatigue.

Complete Counter Balancing

Every possible sequence of conditions or treatments is presented to participants at least once.

Partial Counter Balancing

Altering the order of conditions in order to mitigate order effects. ( A-B-C) and then (A-C-B)

Demand Charactersitics

Exposure to all of the IV may change participant reaction.

Construct Validity

How well the measures were operationalized.

Manipulation checks

An extra dependent variable added to ensure that the manipulation works.

Pilot Study

A simple study used to to confirm the effectivness of the manipulations prior to use.

External Validity

How well the results are generalizable

One Group Design

One group is tested, inflicted, and tested. Bad study.

Maturation Threat

A change in behavior that occurs spontaneously or naturally over time.

History Threat

External or “historical” events that impact most participants at the same time as the treatment.

Regression Threat

People naturally return to baseline, an extreme result in test one is less likely to extreme come test 2.

Attrition Threat

Participants drop out from pretest to posttest, only problematic when systematic

Testing; Order effect

change in participants as a result of expereinceing the D.V. (test) occurring more than ounce.

Instrumentation

When measuring intstruments changes over time.

Placebo Effect

Particpants expectancy of result impacts the result.

Double Blind Study

a rigorous clinical trial design where neither participants nor researchers know who is receiving the active treatment or a placebo, preventing bias in results

Null Result

No difference between groups.

Weak Manipulation

Levels of the I.V. ae too small to induce a response.

Insensitive Measures

Measure cannot detect small differences.

Ceiling Effect

Scores are squeezed together at the high end.

Floor Effect

Scores are squeezed together at the low end.

Manipulation Check

D.V. added to ensure that the manipulation works.

Quasi Experiments

An experimental design in which the I.V cannot be manipulated (pre-existing group)

Quasi-independent Variable

a pre-existing participant characteristic or group difference (e.g., age, gender, culture) that is used to categorize subjects, resembling an independent variable but without direct manipulation or random assignment

Participant Variables

the individual, pre-existing characteristics of research participants

Small - N Design

intensively study a few individuals (often 10 or fewer) through repeated measurements over time, rather than comparing large groups

Behavior - Change Study

intensively study a few individuals (often 10 or fewer) through repeated measurements over time, rather than comparing large groups. Acheivec by establishing a baseline and then comparing baseline to intervention.

Reversal Study

Baseline - Treatment - Removal - Treatmet

Replication

P - Value <.05: 5% chance findings are a fluke, replication allows you to be more sure.

Direct Replication

Reproducing as close to the original study as possible.

Conceptual Replication

The original research question replicated, but different procedures are used. The variables are operationalized differently.

Replication - Plus - Extention

Original study replicated with added variables.

Meta - Analysis

A quantitative technique that creates a mathematical summary of scientific literature on a topic. Mathematical overview. Makes conclusions based off of entire body of evidence.

Review Article

Summarized literature in a narrative ( qualitative ) way.

File Drawer Problem

Null results and opposite results to hypothesis are rarely published.

Underreporting

Researchers mislead results by underreporting conditions or measures that do not support the hypothesis.

P - Hacking

is the misuse of data analysis to find patterns that appear statistically significant when no real effect exists. Researchers may tweak data, exclude outliers, or run multiple analyses until a desired result appears. This practice causes high false-positive rates, compromising scientific integrity. 

Small Sample Sizes

A few chance values can influence data set and results.

HARKing

Hypothesising after results are known.

Ecological Validity

Type of external validity focused on wether a lab study generalizes to real - world situations.

WEIRD Participants

Western, Educated, Industrialized, Rich, Democratic