ch 1 Psychological Science/Research

ch 1

sources of intuition overestimation

hindsight bias

overconfidence

tendency to perceive patterns in random events

Hindsight Bias

Tendency to believe after learning an outcome

“I knew it all along” phenomenon

Overconfidence

tend to think we know more than we do

confirmation bias

tendency to confirm our own preconceptions

Monty Hall Problem

When we go by our gut instead of logic

Game show door (you should switch, but it feels better to stay)

Theory

explanation using an integrated set of principles that organizes observations and predicts the behaviors or events

Hypothesis

Testable prediction often implied by theory

“if…, then…”

operational definition

carefully worded statements of the exact procedure (operations) used in a research study

Replication

Repeating the essence of a research study, usually with different participants in different situations

Research strategies

description, correlation, causation (experiments)

descriptive research

systematic, objective observations of people

goal: provide clear, accurate picture of people’s behaviors, thoughts, and attributes

case study

examines one person or group in depth

BUT doesn’t provide generalized conclusions

ONLY fruitful ideas on the topic at hand

Naturalistic observation

observing a person’s behavior in a natural setting. Only describes the behavior; do not explain

hawthorne effect

behave differently when you know you are being observed (ruins observation)

survey and interview

examines multiple cases in less depth but allows for more generalizations

wording effect

phrasing of questions could bias answers

random sampling

Taking a random group from a population that represents that population.

correlation is NOT … ?

causation (NOT)

correlation defined

research determining if a relationship between two or more factors exists

positive correlation

(0 to +1.00) Indicates a direct relationship where two things increase or decrease together

Negative correlation

(0 to -1.00) Indicates inverse relationship; as one increases, the other decreases

Correlation coefficient

[r] Provides statistical measure of how closely two things vary together and how one predicts the other

example of correlation coefficient

r = -.37

negative weak correlation

variable

anything that can vary and is feasible and ethical to measure

directionality problem

situation where one knows two factors are related but does not know what is the cause and which is the effect

3rd variable possibility

when two correlated variables can actually be explained by a third unaccounted variable

illusory correlation

perception that a relationship between two variables exists when there is little to no relationship

regression toward the mean

tendency for an extreme or unusual score or event but will later fall back to towards the average

experimentation

when researchers focus on possible causes and effects between factors in several ways. Can manipulate the factors of interest to determine the effect

experimental group

participants who are exposed to independent variable

control group

participants who are NOT exposed to independent variable

double-blind procedure

Procedure to eliminate bias in a study. Neither the participants or those collecting data know which group will receive treatment; only the head researcher will know.

placebo

sugar pill

Placebo effect

When people have an expectation that something will bring a certain effect or response

Example of placebo effect

giving someone advil, a well known medicine that helps relieve pain. Patient will immediately exclaim how they feel much better but advil doesn’t work until 20 minutes after consumption. They had the expectation that the drug would make them feel better.

independent variable

factor that is manipulated and being studied

dependant variable

The item is being measured. This changes when the independent variable is manipulated.

confounding variable

factor other than the independent variable that might produce an effect

example of confounding variable

When studying the ability to improve grades with study guides, a confounding variable may be the intelligence of a person and that they do not need study guides to have good grades

random assignment

in order to lessen the effects of a confounding variable in the participants (example: age, gender, beliefs, etc.), researchers can take the participants and mix them up and put each participant into a control or experimental group.

Ethics code of the American Psychological Association (APA) and British Psychological Society (BPS)

1. obtain potential participants informed consent

2. Protect participants from harm/discomfort

3. Keep participant’s information confidential

4. Fully debrief people (explain)

informed consent

giving a potential participant enough information about a study to enable them to choose whether they wish to participate

debriefing

Post experimental explanation of a study, including its purpose and descriptions and deception to participants

mode

most frequently occuring scores in a distribution

mean

arithmetic average of a distribution obtained by adding scores and then dividing by number of scores

median

middle scores in a distribution

range

difference between the highest and lowest scores

positive skewed distribution

when the outlier of data is on the right

negative skewed distribution

when the outlier of data is on the left

standard deviation

computed measure of how much scores vary around the mean

example of standard deviation

If the IQ of a group’s standard deviation is 10 units, then most of the group will have an IQ about 10 units from the mean

small Standard Deviation

scores a concentrated and less scattered

variants

average of the squared difference from the mean

rad(variants) = Standard deviation

normal curve distribution

symmetrical, bell-shaped curve that described the distribution of data; most scores are near the mean with very few extremes

(68% fall within one standard deviation)

Normal distribution curve example: exam scores

mean = 80

standard deviation = 90-80 = 10

z scores

the number of standard deviations a value is from the mean of a given distribution.

normal distribution percentages

68%, 95%, 99.7%

When is an observed difference reliable?

1. people are proper representatives

2. small standard deviation (concentrated)

3. More cases can allow for generalization

Generalizations based on a few unrepresentative cases are ___________

unreliable

When is an observed difference significant?

when the sample averages are reliable and difference between them is relatively large

(experimental and control results contrast a lot)

statistical significance

likelihood that results happen by chance;

statistical statement of how likely it is that an obtained results occurred by chance

research beyond a reasonable doubt means that the odds of an outcome’s occurrence by chance are less than ________

5% or .05

Example of statistical significance

When testing a new speech therapy to an established technique (control), the p-value was 0.032. This statistical statement shows how likely results occurred by chance.

statistical significance does not equal

practical significance

what is practical significance

the likelihood of something to occur in real life.

Example of stat. sig. not being practical sig.

Study on aspirin: It can lower the chances of heart attacks (statistically proven). But people still got heart attacks. Just because the numbers say taking aspirin will lower chances of heart attack, other factors play a role like sleep, exercise, and minimizing stress.

Type 1 error

concluding a difference when there is none (false positive)

Type 2 error

conclude no difference when there is one (false negative)

example of type 1 error

when taking a pregnancy test, a couple believes they are pregnant when they are actually not

example of type 2 error

After a COVID test, it states that a person does not have COVID when they actually do.