Exam 3 Research Methods

What is positive correlation?

A relationship where two variables move in the same directions (in a graph they both go up or both go down)

What is negative correlation?

A relationship where two variables move in opposite directions (in a graph they both go in different directions so one goes up and one goes down)

Why doesn’t correlation prove causation?

Due to third variables, coincidence, or directionality problems that can influence and explain the relationship.

Third variable problem

A hidden variable causes both variables being studied. (Ex. People who carry lighters may have higher rates of lung cancer - doesn’t mean lighters cause cancer, the third variable is smoking, as smoking causes both lung cancer and carrying lighters)

Directionality problem

Researchers cannot determine which variable causes the other. (ex. Stress and lack of sleep, does stress cause lack of sleep or vice versa?)

Stroop Effect

Interference caused when automatic processing conflicts with controlled processing

What is automatic processing?

Fast, unconscious processing that requires little effort. (Ex. Reading)

What is controlled processing?

Slower processing that requires attention and conscious effort. (Ex. Naming colors)

What is interference in the Stroop task?

When two stimuli conflict (Ex. The word RED in the color green)

What is facilitation in the Stroop task?

The stimuli match/are the same (Ex. The word RED in the ink color red)

Small-N Designs

Very small studies (1-10 people) that follow the participants over time, and are good for things that don’t impact larger populations. (Ex. Studying specific rare diseases)

What are operant schedules and why do they work well with small-N designs?

Operant schedules are patterns of reinforcement that determine when a behavior is rewarded. They work well with small-N designs because researchers can closely track individual behavior changes over time and clearly see how reinforcement affects behavior.

What is A Baseline?

The original level of behavior before treatment.

What is B Baseline?

Treatment, introducing treatment and looking for changes.

What is ABA design?

Reversal design, removing the treatment and seeing if it will be successful by going back to original A baseline. (Helps show cause and effect)

Why use multiple baselines design?

When removing treatment is unethical or impractical (ex. Learning a skill), you then stagger the treatment by using it at different times, with different people or in different situations.

What is operant conditioning?

Learning through rewards and punishments

What is positive reinforcement?

The addition of something pleasant to encourage behavior. (Ex. A treat)

What is negative reinforcement?

The removal of something unpleasant to encourage behavior. (Ex. Shock collar that stops when barking stops)

What is positive punishment?

Adding something unpleasant to decrease behavior. (Ex. Spank on the butt)

What is negative punishment?

Removing something pleasant to decrease behavior. (Ex. Ignoring a dog when it barks instead of giving it attention that it wants)

What is shaping?

Rewarding progress (Ex. Giving a treat for every small step towards the final goal)

What is internal validity?

Confidence that the independent variable caused the observed changes.

Key threats to internal validity?

Maturation & History

What is maturation?

Subjects changing naturally over time

What is history?

External events influencing the outcome

How can you improve internal validity?

By using a nonequivalent control group

What is a nonequivalent control group?

A group used in a study that is similar to the treatment group but not identical and not randomly assigned. (Ex. One school gets new curriculum, and a similar school keeps the old curriculum)

What is a subject variable?

A characteristic that participants already have that cannot be changed and that researchers cannot manipulate. (Ex. Sex, age, ethnicity)

What is matching?

Paring people different groups (treatment vs control) so they are similar on important characteristics before comparing them.which can help improve internal validity. (Ex. For every student in treatment group with a certain IQ there is a student in the control group with a similar IQ)

What are some problems with matching?

You aren’t able to account for every interaction (it doesn’t eliminate compiled but it can reduce compounds)

What are compounds?

More than one thing happening together, so you can’t tell which one is causing the effect.

What is a cross-sectional design?

Comparing different age groups at one time, group A compared to group B. (Ex. Studying 15 year olds vs 12 year olds, different people at the same time)

What is a time-lag design?

Following the same participants over long periods of time. (Ex. Studying 12 year olds until they turn 15)

Problems with longitudinal design?

Outside influences that cause things to change.

What is a time-lag design?

Comparing same-age participants at different times. (Ex. Studying 50 year olds in 2000, 2005, 2010, and 2020)

What is a cross-sequential design?

Mix design, studying different age groups over long periods of time. (Ex. Studying 10, 15, and 20 year olds and testing them now, in 2 years, and in 5 years)

Problems with time lag design?

Cohort effects (different generations, upbringings, life experiences, etc) & history (changes in historical events and life)

What is a quasi experiment?

An experiment without random assignment. The only thing that makes it quasi is that the researcher is not directly manipulating the independent variable.

What is natural treatment?

A type of quasi experiment where the actual “treatment” happens naturally , and you study how behavior changes when the treatment/event happens in the real world and not in a controlled lab experiment. (Ex. Introducing a new school curriculum)

Problems with natural treatment?

No reversal (can’t remove things to confirm causality, ex. Grades improved but you don’t know if it’s from the new curriculum), maturation (people naturally changing overtime, ex. Students getting older, confident, more experienced and improving due to that and not the new curriculum), history (outside events that can effect results, ex. Test scores improve be due the school started after school tutoring at the same time)

What is a ceiling effect?

Scores cluster near the top because the test is too easy.

What is a floor effect?

Scores cluster near the bottom because the test is too difficult.

What is scale attenuation?

When the measurement scale being used is too limited making it difficult to see the relationship between the variables because the scores are too similar. (Happens only at the bottom and top)

What is interaction?

Treatment works differently for different groups (ex. A study program is given, low performing students improve while high performing students do not)

What is regression?

Extreme scores move closer to average over time. (Ex. Student was sick and scores a 10/100 on a test, then scores a 50/100 on the next test, looks like improvement but they just moved closer to the average)

What is replication?

Repeating a study to see if the results hold up

What is direct replication?

Replicating the study in the exact same way as close as possible (to make sure there was no messups)

What is systematic replication?

Repeating the study but changing minor factors (to see if it’s reliable and general)

What is conceptual replication?

Testing the same theory using different methods. (To test the theory)

Electroencephalography (EEG Machine)

Measures electrical activity in the brain using sensors on the scalp in order to help understand the brains activity.

Magnetic Resonance Imaging (MRI Machine)

A machine that uses strong radios and magnets to take detailed picture of the brain so we can better understand the brains structure.

Perfect correction between two variables means what?

High degree of validity and linear relationship.