Neur Research midterm 2

what is the primary goal of the experimental research strategy

to establish a cause and effect relationship between variables. Changing the independent variable (IV) causes a change in the dependent variable (DV)

What are the four basic elements of an experimental research strategy?

1. Manipulation

2. Measurement

3. Comparison

4. Control.

What is the purpose of Manipulation in an experiment?

To manipulate the independent variable (IV), which helps determine which variable is the cause and which is the effect.

What is the purpose of Measurement in an experiment?

To measure the dependent variable (DV) and obtain scores for each treatment condition

What is the purpose of Comparison in an experiment?

To compare the scores obtained in one condition with the scores obtained in another condition

What is the purpose of Control in an experiment?

To control all other variables so they do not influence the two variables being examined - involves eliminating or controlling confounding variables

What is a Confounding Variable?

A variable that changes with the independent variable and can affect the dependent variable.

what happens to a obscuring variable when it influences the dependent variable (the result)

it becomes confounding

What is an Obscuring Variable?

A variable that makes changes in the dependent variable hard to observe

what do obscuring variables lead to

Lead to measurement error and noisy data, reducing internal validity.

How can researchers control obscuring variables?

1. By ensuring the manipulation is effective (adequate strength)

2. reducing measurement error (e.g., improving instrumentation or training)

3. and minimizing excessive variation in data (often due to individual differences)

What is the Directionality Problem?

A challenge that occurs when a research study establishes a relationship between two variables, but does not explain the direction of the relationship (which one causes the other)

What are the two critical components of an experiment?

The Independent vs. Dependent variable and the Experimental vs. Control group

In experimental terminology, what is a Factorial design?

A design that includes more than one Independent Variable (IV)

In experimental terminology, what is a Multivariate design?

A design that includes more than one Dependent Variable (DV)

List three methods used for control in experiments.

Holding variables constant (easy for things like light, temperature, noise); Matching; and Randomization

What is the downside of holding variables constant or matching for control?

t may limit generality (external validity)

What is a Control group?

The group NOT exposed to the manipulation

What is a No-Treatment control group?

A control group where subjects do not receive the treatment being evaluated

What is a Placebo control group?

A control group where participants receive a placebo (e.g., a sugar pill) instead of the actual treatment. This is typically used in clinical studies

Define the Placebo effect.

Responses to the fake medication, meaning people experience improvement even though they received a dummy sugar pill. This effect involves the power of expectation and care

Define the Nocebo effect.

Experiencing negative symptoms due to expected side effects

How does the experimental research strategy typically balance validity?

It allows careful control of the environment, leading to high internal validity, but often uses unfamiliar environments, which limits external validity

What is Internal Validity?

The confidence that the research produces an unambiguous explanation for the relationship between two variables, demonstrating that change in the DV MUST be due to change in the IV within the experiment.

What is External Validity?

The extent to which your experimental result holds true outside your specific study. This involves generalizing results to different samples, settings, or measurements

What is a major threat to internal validity?

Any factor that raises doubts or allows alternative explanations for the relationship between variables (ex. assignment bias, environmental variables, and time-related variables)

List the three general categories of threats to external validity.

1. Generalizing across participants or subjects

2. Generalizing across study procedures

3. Generalizing across different measures.

List examples of threats to external validity related to participants/subjects.

• Selection bias (favoring certain individuals)

• use of a convenience sample (which can be biased)

• volunteer bias (volunteers differing from non-volunteers)

• specific participant characteristics (demographics)

• cross-species generalization.

List examples of threats to external validity related to study procedures.

• Novelty effects (participants influenced by being in a new environment)

• Multiple treatment interference (effect from one trial carrying over to the next, like fatigue or practice)

• Experimenter characteristics (demographics of the researcher affecting responses)

List examples of threats to external validity related to measures.

Sensitization (monitoring a behavior changes the behavior), and the Timing of measurement (when the measure is administered)

What is the fundamental difference between Between-subjects design and Within-subjects design?

In Between-subjects design, subjects are randomly assigned to groups, and each group gets a different treatment.

In Within-subjects design, all subjects participate in both groups/treatment levels (repeated measures design).

What is the primary advantage of the Between-subjects design?

Each measured score is independent of the others. Results are less influenced by factors like fatigue or practice from multiple testing/exposure to other treatments

What are the disadvantages of the Between-subjects design?

1. Requires more individuals

2. Large differences between individuals in each group can obscure (hide) differences between the groups (due to high within-group variability)

What is Variability?

Variability refers to the distribution of scores

What is Variance?

The statistical value for the difference of each value from the mean. Large differences between individuals produce large variance 

In a Between-subjects design, what is the goal regarding variability?

To Minimize within-group variability and maximize between-group variability

What does Between-group variability refer to?

Differences between treatments. Researchers hope to see these differences

What does Within-group variability refer to?

Differences within a single treatment group. It is unavoidable but should ideally be kept equal across groups

How can researchers mitigate variance in a Between-subjects design?

by using standardize procedures:use a very large sample size (a statistical feature that overcomes negative effects of high variance); and ensure group equivalency (e.g., using random assignment with matching or restricting the range of baseline characteristics in animals)

Why is random assignment not sufficient to ensure groups are identical in a Between-subjects design?

While random assignment ensures there are no systematic differences, it does not guarantee the groups are perfectly identical, meaning individual differences (e.g., age, weight) might still lead to initial differences between groups before manipulation

What are the two designs used to control for the nonequivalence of groups in a Between-subjects experiment?

Pretest-posttest design and Matched pairs design.

Describe the Pretest-posttest design (Between-subjects).

Groups are measured before manipulating the IV to check for equivalence, and then measured after. The goal is to compare the treatment vs. control group results (not pre-test vs. post-test).

What is the main advantage and disadvantage of the Pretest-posttest design?

Pro: It tests the equivalence of groups and limits threats to internal validity.

Con: It is time-consuming and may sensitize participants to the procedure

Describe the Matched pairs design (Between-subjects).

Participants are grouped into pairs based on important characteristics/greatest similarity during random assignment, and then one member of the pair is assigned to each condition (treatment/control)

What is the main advantage and disadvantage of the Matched pairs design?

Pro: Minimizes the effect of individual differences.

Con: Attrition (if one matched partner drops out, the other must be removed too), and it can be difficult to match people exactly

What is the primary advantage of the Within-subjects design?

Fewer participants are needed, and it eliminates variability due to individual differences between treatment groups, as the same individual is in every condition

List the four categories of time-related effects that threaten the Within-subjects design.

1. History (outside events affect results)

2. Maturation (physiological/psychological changes over time)

3. Instrumentation (changes to the measuring instrument)

4. Order effects (Practice/Fatigue).

Define the two components of Order effects (a time-related threat).

• Practice: Performance improves with repeated measures.

• Fatigue: Performance deteriorates with repeated measures.

What are Sequence effects?

An interaction between each condition in the experiment that reduces the sensitivity of the response, potentially causing the participant to become sensitized or desensitised

What is the solution used to control some time-related effects (specifically Order effects) in Within-subjects design?

Counterbalancing

What is the goal of Counterbalancing?

To use every possible order of treatments with an equal number of individuals participating in each sequence. This balances the order effect equally among treatment groups, minimizing the chance that practice/fatigue becomes a confoun

How does a quasi-experimental or nonexperimental strategy differ from an experimental strategy regarding group assignment?

Experimental designs use random assignment;

quasi- and nonexperimental designs use pre-existing groups or variables since the researcher can’t manipulate or randomly assign the independent variable.

What is the main difference between Nonexperimental and Quasi-experimental strategies?

Quasi-experimental designs include some control over internal validity (like comparison groups or pretests), while nonexperimental designs have no control at all.

What is the major problem posed by using a pre-existing independent variable?

Using a pre-existing variable prevents random assignment, so results may reflect confounding differences between groups rather than the independent variable — this is called assignment bias.

What are Transgenic mouse models?

Transgenic mice have artificial DNA (transgenes) inserted into their genome to study an independent variable — either knockin (gene added/overexpressed) or knockout (gene deleted/inactivated).

Why do transgenic mouse models complicate interpretation of results?

Because the mouse can develop compensatory changes to adapt to the gene alteration, results reflect the net effect of multiple variables, not just the transgene itself.

In a Transgenic model, what are Littermate controls?

Mice generated in the same litter that do not contain the experimental transgene, used to eliminate contributing effects from the mother (dam).

In quasi/nonexperimental notation, what do 'X' and 'O' represent?

X = intervention, treatment. On = observation at time n.

What is Assignment bias in the context of nonequivalent group design?

The assignment of individuals is not controlled, resulting in groups that are noticeably different. It is a built-in threat to internal validity.

Describe the Differential research design (Nonexperimental).

Differential design compares pre-existing groups (e.g., men vs. women) without manipulation to see if a difference naturally exists between them.

Describe the Posttest-only nonequivalent control group design (Nonexperimental).

Nonequivalent control group design: uses two pre-existing groups — one receives the treatment (X), the other does not — and compares their post-treatment outcomes (O).

Describe the Pretest-posttest nonequivalent control group design (Quasi-experimental).

Pretest-posttest nonequivalent control group design: measures both groups before and after treatment (O X O); the pretest helps control assignment bias, making it quasi-experimental.

What is the difference between Time-related designs and standard Within-subjects designs?

Time-related designs observe the same group over time like within-subjects designs, but can’t use counterbalancing, since the pretest always comes before the posttest, leaving order effects uncontrolled.

Describe the One group pretest-posttest design (Nonexperimental).

One-group pretest-posttest design: measures one group before and after treatment (O X O) with no control group.

Describe the Time series design (Quasi-experimental).

Takes multiple measurements before treatment and multiple measurements after treatment.This improves confidence by checking for pre-existing trends and permanence of the effect

What is an Interrupted time series design?

A time series design where the treatment or manipulation (X) is naturally occurring and not induced by the researcher.

What two developmental research strategies study change as a function of age?

Longitudinal design (within-subject) and Cross-sectional design (between-subject)

Describe the Cross-sectional design (Between-subject).

Studies individuals of different ages at the same point in time.

Con: Vulnerable to cohort effects (differences between age groups due to unique characteristics of their generation).

Describe the Longitudinal design (Within-subject).

follows the same individuals over time; pro: tracks changes within the same people; con: time-consuming, costly, high dropout rates.

In general, how do quasi-experiments rank in terms of validity?

usually have high external validity (realistic settings) but lower internal validity (confounds weaken causal claims).

Define the term Population in research.

The entire set of individuals you are interested in that share some common characteristic

Define the term Sample in research.

A set of individuals that represent the population; a subset of the population

What is a Representative sample, and why is it crucial?

A sample that has the same characteristics as the population. Generalizing results from the sample to the population is accurate only when the sample is representative

What is a Biased sample?

A sample with different characteristics from those in the population.

What is Selection bias?

A sampling procedure that favors the selection of some individuals over others

What are the two primary methods for determining sample size?

Power analysis and the Resource equation.

What is Power analysis used for?

To calculate the minimum sample size required.

What are the four major factors influencing sample size calculation via Power analysis?

1. Size of the difference you need to detect (Effect size)

2. Variability in the factor of interest (Standard deviation)

3. P value planned for statistical significance (Type 1 error, typically 0.05)

4. Power (how confident you want to be that you will detect statistical significance, usually 80%)

When is a larger sample size needed?

If the difference between groups is small (decreases) or if the standard deviation is large (increases)

When is the Resource equation typically used instead of Power analysis?

1. When it is not possible to predict an effect size (mean)

2. standard deviation (no previous findings available),

3. or when multiple endpoints or complex statistical procedures are use

State the Resource equation formula.

E = (Total number of animals) − (Total number of groups)

What range should 'E' fall within for the sample size to be considered adequate using the Resource equation?

E must be between 10 and 20

What is the main requirement for Probability sampling methods?

The researcher must know the exact size of the population and be able to list everyone in that population

What are the four general methods of Probability sampling?

Simple random, Systematic, Stratified, Cluster

Define Simple random sampling.

A method where all individuals have an equal chance of being selected, and each selection is independent of the others.

Differentiate between Sampling with replacement and Sampling without replacement.

With replacement: chosen individuals go back into the pool and can be selected again.

Without replacement: chosen individuals are removed, changing probabilities but ensuring no repeats.

Define Systematic sampling.

A method where the researcher randomly selects the first participant, and then selects every nth participant on a list thereafter

Define Stratified random sampling.

A method where the population is divided into strata (subgroups), and then participants are selected from each stratum

Differentiate between Simple stratified sampling and Proportionate stratified sampling.

• Simple stratified: Obtains equal numbers from each strata.

• Proportionate stratified: Selects a different proportion (%) from each strata, based on the actual representation in the population

Define Cluster sampling.

A method where the population is identified into clusters (pre-existing groups), and the researcher chooses from a proportion of the clusters.

What is the main difference between Non-probability sampling and Probability sampling regarding the population?

In Non-probability sampling, the population is typically not completely known, so the exact probability of selecting any individual is unknown or difficult to determine

What are the two general methods of Non-probability sampling?

Convenience and Quota

Define Convenience sampling.

Convenience sampling: selects participants who are easy or available to access; common but inherently biased.

Define Quota sampling.

A method used to identify a subgroup and then choose individuals from that subgroup to reach a required quota

How does Simple quota sampling differ from Simple stratified sampling?

Simple stratified sampling: randomly selects individuals from each subgroup (probability).
Simple quota sampling: fills subgroup quotas without random selection (convenience, non-probability).

Define Proportionate quota sampling.

Proportionate quota sampling: selects participants so subgroup proportions match the population (e.g., 60% female, 40% male).