AP psych stats

Hypothesis

Statement that is rooted in theory; can be falsifiable

Falsifiable

Can be proven to be incorrect

Qualitative research

Non-mathematical tactics of research

Quantitative research

Mathematical tactics of research; using a tool

Population

All people

Sample pool

People that are available to be a part of the study

Representation

How much the sample reflects a populations demographics

Generalizability

How well data can be applied to a larger demographic

Sample

The specific people chosen from the sample pool

Variables

Things that are being measured/evaluated

Random sample

People that have an equal chance of being chosen to participate in a study

• Convenience sample

Participants that are easily available to the experimenter (prevents sample bias)

Replication

Repeating experiment to increase confidence in results

Validity

Applies to the TOOL/TEST - accuracy; does it measure what its intended to?

Reliability

Replication of results (data) - precision; are the results consistent as you repeat it?

Confederate

Someone that knows about the researchers experiment, that is pretending to be a part of the study

Research types:

Experimental study

Cause and effect; experimental manipulation of a variable(s) to see how they impact another variable(s)

• Causation

Cause an effect; how one variable effects the outcome of another

Correlation study

NOT causation; NATURALLY occurring relationship between two variables

• Correlation Coefficient

Measures the direction and how close these two variables are ± 1

Descriptive study

Describing behavior through case studies, surveys, naturalistic observations —> BASIC info

Research design methods:

Survey

Questions —> can lead to self reporting bias; when you choose what information to include/exclude

Naturalistic observation

Observing peoples behavior in their natural environment without any interference

Case study

Observes the behavior(s) of a singular person or a small group of people; a form of longitudinal study; used for rare conditions

Research design techniques:

Single-blind procedure

When the participant is blind to the group that they were placed in a study

Double-blind procedure

When BOTH the experimenter and the participant are blind to who is in what groups in the study

Longitudinal study

SAME group of people are involved in a study

Cross-sectional study

Population of people studied at the same time and the researcher is comparing different sections of that population with each other; ex. Maria wanted to see if there is a relationship between age and willingness to try new things. She studies 5 year olds, 10 year olds, 15 year olds, and 20 year olds —> different demographics regarding the same topic

Experimental design elements:

Placebo

Inactive (inert) substance that has no effect and given to the control group; prevents participant bias (they don’t know if they’re in the control group or experimenter group

Control group

Group that does not recieve IV - comparison purposes

Experimental group

Group that receives IV

Operational definition

States what the the variable means and how it will be measured (TOOL & STATISTIC IS A MUST)

Random assignment

Putting participants into either the control group or experimental group by chance

Independent variable

Variable manipulated by researcher — experiments

Dependent variable

Variable that is affected by the independent variable

Constant variable

Variable that is kept the same throughout the entirety of the experiment

Problems:

Confounding variables

Outside factors that influence the outcome of the exp; something other than the IV affects the DV

Third variable problem

Other factor that is affecting study outcomes

Sample bias

Sample CANNOT be applied to the general population

Participant bias

Participants alter their behaviors in an experiment

• Social desirability bias

Changing the way that you present yourself to fit the goals of the researchers experiment or for privacy purposes

• Hawthorne effect

Changing your behavior because you know that you are being watched

• Self-reporting bias

Choosing what information to include/exclude (ex. In a survey)

Experimenter/Researcher bias

Experimenter tries to make data fit in with their study goals

Ethical guidelines:

Informed consent

Giving background about what the experiment is about and their rights

• Informed assent

For children that are under the age of 18; parents consent but the child also has to a fill a form out to ensure that they really want to do this study

• Right to withdraw

Participants have the right to leave a study whenever they feel that it is best for them to; should be in the informed consent form

• Deception

If someone is decepted in a study, the experimenter MUST debrief them, allow them to follow up with any questions, and provide any aftercare if need be

• Debrief

Experimenter tells decepted participants their true intention for their study

Confidentiality

All participant information shall be private and protected

Anonymity

Participant data should be kept anonymous

Long-term harm

Aftercare needs to be provided

Animal research

Do not harm animals for no reason

Institutional review board (IRB)

Approves research regarding humans

Peer review

Research/work is reviewed my licensed researchers

Descriptive Statistics:

Measures of central tendency

Mean: sum of cases divided by total of cases

Median: splits our data in half - ALWAYS put in order FIRST

Mode: number that repeats the most in a data set

Measures of central variability

Range: HIGHEST - LOWEST

SD - Standard Deviation: average difference in a data set

Frequency

How often something is appearing

Percentage

Ratio of the WHOLE - segments between SD

Percentile

Ratio of scores falling BELOW that point - ex. 78th percentile

Correlational coefficient

Measures the direction and strength between two variables

Positive correlation

As one increases, the other increases

Negative correlation

As one increases, the other decreases

Illusory correlation

The belief that there is a relationship between two variables (even though there isn’t!)

Distribution

Skewed distribution: use the median —> if you use the mean, the average would not correctly reflect all of the participants since its taking into account all the outliers in the data set

Bi-modal distribution: use the median (same reason as above)

Normal distribution: use the mean

Normal curve

Mean is used; low variability (majority of data is near the mean)

Mean median and mode are all located at the same place (in the center)

Bimodal distribution

Two separate curves; mean in the middle; has two modes

Skewness

Where the curve is located; graph that has a few extreme scores that are forming an asymmetrical curve

Higher variability, the wider the graph

Lower variability, the skinnier the graph

• Positive skew

Mean is higher than the median

• Negative

Mean is less than the median

Inferential statistic:

Statistical significance

Other stats:

Effect size

How much difference between groups; the p > 0.05 Or p < 0.05

If p is < 0.05, results are most likely NOT due to chance

If p is > 0.05, results are most likely due to chance

Regression towards the mean

Extreme scores gradually move towards the mean when study is repeated; can happen because random factors that were affecting the outcome of the study are no longer present

Extra info

When a question asks what is the percentage of kids above/below a certain number in a bell curve (use the 68%)

If a question asks between an SD interval, USE the 34.1%

Remember the 13.6 34.1 34.1 and 13.6 for a NORMAL distribution

Scatterplot

Dots; can have a strong/weak/negative/positive relationship; all depends on the direction of the dots and the distances that they are from one another

Line of best fit

A Line that best demonstrates the results of a specific scatterplot (cc of 0 is just a horizontal line)

Bar graph

Categorical data

Histogram

The bell curve - quantitative data

Line graph

Linear relationship