Study Guide Chapter 11: Internal Validity and Experimental Design

These flashcards cover key concepts related to internal validity and experimental design from Chapter 11 of the study guide, focusing on threats, effects, and research methodologies.

What are threats to internal validity?

Any alternative explanation for an observed relationship between the independent and dependent variables. (design confounds, selection effects, order effects)

What is a comparison group used for?

To determine whether changes in the dependent variable were caused by the treatment, a threat to internal validity, or both.

helps rule out many threats to internal validity

what is maturation

An internal change in participants over time, like growing, healing, or getting tired.

the participants change internally

to fix you need to add a comparison group

What is a history effect?

An external event that occurs between pretest and posttest that causes a change in the dependent variable.

• To fix?  have a comparison group that experiences the same outside event but not the treatment

What does 'regression to the mean' mean?

When an initial score is extremely high or low, the second measurement is likely to be closer to the average, even without any treatment.

• if the first measure is extreme, then the second measure will likely be closer to the average

What is the definition of regression?

The tendency for the second measure of the dependent variable to be closer to 'average' than the first measure.

What is attrition?

When participants drop out before the posttest, possibly leading to biased results if certain types of people drop out.

• What changes? Sample shrinks

• Ex. only motivated people return

• Fix it? Remove pretest data of those who dropped out

How can a history threat be fixed?

Have a comparison group that experiences the same outside event but not the treatment.

What is a testing effect?

When taking a pre-test influences the post-test, causing improved performance due to familiarity with the test.

What is instrumentation effect?

When the measurement tool changes between pre-tests and post-tests, leading to a false change in the dependent variable.

Describe a one-group pretest-posttest design and its drawbacks.

Design: 1. Measure the DV (pretest), 2. Give treatment, 3. Measure DV again (posttest). It is bad because it contains no comparison group, making is vulnerable to time-related threats: maturation, history, regression to the mean, attrition, testing, and instrumentation. 

It's a design with no comparison group, making it vulnerable to various time-related threats.

What is a confound?

A variable that influences both the independent and dependent variables, creating an alternative explanation.

How should confounds be addressed?

Using random assignment during the study design.

What are demand characteristics?

Cues that lead participants to guess the hypothesis and change their behavior accordingly.

What is a single-blind study?

A study where researchers do not know which condition participants are in.

What is a double-blind study?

A study where neither the participants nor the experimenters know which condition participants are in.

What is observer bias?

When observers’ expectations influence how they record or interpret behavior.

What is the placebo effect?

Improvement in participants due to the belief they are receiving an effective treatment, even if it is inactive.

What are design confounds?

A type of confound where a variable systematically varies with the independent variable due to poor study design.

What does 'post hoc ergo propter hoc' mean?

“After this, therefore, because of this.”

Assuming that because B happened after A, A must have caused B; a logical fallacy.

What is the 'Law of Similars' in homeopathy?

LIKE CURES LIKE

The idea that a substance causing symptoms in a healthy person will cure those same symptoms in a sick person.

What is the 'Law of Infinitesimals'?

The principle that diluting a substance increases its potency and curative power. Stems from a belief that large doses of a substance can worsen symptoms, but diluted into a very small amount can have a healing effect.

What is the first step in Freirich’s experimental plan?

Pick any treatment appropriate for the condition being studied. PICK AN TREATMENT

What logical error does 'post hoc ergo propter hoc' represent?

A false causal inference assuming effectiveness simply because improvement follows treatment.

When do you apply treatment according to Freirich's experimental plan?

Apply treatment after observing progressive worsening of the patient's condition.

What conclusion can be drawn if the patient approves after treatment?

It works or was beneficial.

What conclusion can be drawn if the patient worsens after treatment?

the dosage needs to be increased or they need more time

What should be done with the dosage if the patient improves?

The dosage should be decreased.

Why does prior plausibility matter?

Extraordinary claims require extraordinary evidence to be accepted.Because the more a claim contradicts established knowledge, the stronger the evidence must be to support it 

How does regression to the mean apply to Freirich's experimental plan?

Patients often seek treatment at their worst symptoms, and improvement may be due to regression to the mean rather than treatment effects.