Research Psychology Exam 4

What are the three principles of graphing?

Clarity, Precision, and Efficiency

Clarity

If you have a story to tell, make sure it is easy to understand, representing data in a way that is
closely integrated with the numerical meaning

Precision

If you have a story to tell, make sure it is presented accurately. represent numbers with exactness, no exaggerating

Efficiency

If you have a story to tell, get to the point

Frequency Distribution

Allow you to present an overall pattern of data, as a table or graph

The dependent variable is almost always

on the y-axis (ordinate)

Central Tendency

“what is typical”

Mean

The average or score

Median

the score in which 50% of of the data is higher and 50% is lower

Mode

Most common score

Range

The difference between the lowest and the highest observed scores

Extended Range

The difference between the lowest and highest observed scores plus 1

The larger the sample (N) is

the more likely it is to be unbiased

The mean score is

a point estimate of the true score of the population

Interval Estimate is

a measure of how certain one is in the point estimate

A common interval estimate

is the 95% confidence interval

Standard Normal Distribution

Normal, bell-shaped distribution with a mean value of 0 and standard deviation of 1

Z scores

standard score, units of standard deviations, tell you how far from the mean score an individual score is

How do we know causal inference? What causes it what?

Co-variation, temporal precedence, internal validity

Co-variation

Are variables X and Y mutually related?

Temporal Precedence

Y must not occur until after X occurs

Internal Validity

Can the relationship between X and Y be attributed to some other variable?

The Third Variable Problem

X and Y might both be related to a third variable Z.

Continuous Variables

may take on infinite number of values (real numbers, GPA)

Discrete Variable

can only take a limited number of variables (integers) ex. grades

Dichotomous Variables

may only take on two values (true or false) ex. gender

Correlation Coefficients measure

the strength to which different variables are related

Pearson r

Two continuous variables, like correlation between SAT scores and GPA

Spearman rho (Rs)

two ranked variables

Point biserial (Rpb)

One continuous and one dichotomous variable, correlation between gender and with SAT score

Phi Coefficient

two dichotomous variables, correlation between gender and yes-no answers to a survey question

Rxy

Average product of z scores of X and Y

What is the Pearson r measure called

product (Zx times Zy) - moment (z scores are the distance from the mean) correlation

Life Stress

when external changes occur in a person’s life and the individual needs to make psychological adjustments

Holmes and Rahe

their goal was to develop a way to measure the size and magnitude of stressful life events

When the difference score (D) is positive

it says the first set of rankings is less extreme than the second set

Spearman Rho is called a

product moment score

Dummy Coding

changing dichotomous variables to 0 and 1s

NHST

Null Hypothesis Significance Testing, finding the probability value that the observed difference between two means is not chance

Null Hypothesis (H0)

A certain variable has no effect on a dependent variable. Ex. There is no difference between A and B items.

Alternative Hypothesis (Ha)

A certain variable does have an effect on a dependent variable. Ex. A items are larger than B items on average.

What is the probability that the observed difference between A and B items is due to chance?

p value

If p is less than 0.05

you reject the null hypothesis

Type 1 Error

The error of rejecting the null hypothesis when it is actually true.

The probability of a Type 1 error is determined by

the significance level from a statistical test

Type 2 Error

The error of failing to reject the null hypothesis when it is false.

The probability of avoiding a type 2 error is

the statistical power of the test

Statistical Power

The probability a significance test will reject the null hypothesis when it is false

If r is the product moment statistic of correlation

r tells us about the robustness of the relationship between two variables

Significance Test (p) =

effect size x size of study

A large effect size might not be statistically significant if

it is not based on many observations

A small effect size can be significant if

there is a high-level power in the design

How are bar graphs and line graphs used to visualize data?

Bar graphs display discrete categories with rectangular bars, while line graphs show trends over time or continuous data using points connected by lines.

How do stem-and-leaf charts work?

They organize data to show distribution; each value is split into a "stem" (e.g., tens) and a "leaf" (e.g., ones), maintaining original data values while showing frequency.

When are means, medians, and modes used?

Mean is for symmetrical data; median is for skewed data; mode is for categorical or multimodal data.

How is a confidence interval around a population mean computed?

CI = sample mean ± (critical value × standard error); shows range where the population mean is likely to fall.

Which range should be reported?

Report the interquartile range (IQR) for skewed data and the full range (max - min) for a general spread.

What is the difference between variance and standard deviation?

Variance is the average squared deviation from the mean; standard deviation is the square root of the variance.

What is the purpose of descriptive and inferential measures?

Descriptive stats summarize data; inferential stats make predictions or test hypotheses about populations using samples.

What is the role of the standard normal distribution?

It allows comparison across datasets by standardizing values, assuming a mean of 0 and standard deviation of 1.

Why are z-scores called standard scores, and how are they used?

Z-scores standardize data by showing how many SDs a value is from the mean; used for comparison and identifying outliers.

How are correlations visualized in scatter plots?

As dots representing paired values; patterns show strength/direction of relationships (linear, curvilinear, or no correlation).

How is a product moment correlation calculated?

Using Pearson’s r: a formula comparing covariance of variables to the product of their standard deviations.

Why is it important to focus not just on statistical significance?

Because significance doesn’t indicate effect size or practical importance—small effects can be significant with large samples.

What is the distinction between Type I and Type II errors?

Type I: Rejecting H₀ when it’s true (false positive). Type II: Failing to reject H₀ when it’s false (false negative).

What are one-tailed and two-tailed p values?

One-tailed tests look for an effect in one direction; two-tailed tests check for effects in both directions.

What does statistical power have to do with NHST?

Power is the probability of correctly rejecting a false null hypothesis; influenced by effect size, sample size, and alpha level.

What is effect size (r)?

A measure of the strength of a relationship or difference; r indicates correlation strength (small ~.10, medium ~.30, large ~.50).

Abscissa

horizontal x axis

Ordinate

Y axis

Line Graph

Efficient way to graph changes in frequency of scores over time

Stem and Leaf Chart

Efficient for presenting a batch of data

Exploratory Data Analysis

looking for clues, identifying patterns in data

Trimmed Mean

Researchers prefer a trimmed mean to account for extreme outliers

Variance

Tells use deviation from the mean of scores, referred to as mean square

Standard Deviation

square root of population variance

descriptive measure

the formula for measuring variability describes a complete population of scores and events

Inferential Measure

the equation used to measure variability of a sample of scores

S²

Unbiased estimator of the population value

In a symmetrical distribution,

median and mean have the same value

We calculate standard z scores by

transforming raw scores into standard deviation units

beta

probability of making a type 2 error

Error of gullibility

making type 1 error

Blindness

making type 2 error

2 tailed p value

Alternative hypothesis didn’t specifically predict which side of the probability distribution the significance will be detected

One tailed p value

applicable when the alternative hypothesis requires significance to be in one tail or the other

counternull statistic

• It shows that a small effect with p = 0.05 doesn't mean the effect is definitely real or meaningful.

• It helps researchers visualize the uncertainty around their results, especially when p-values are borderline.

Statistical Power

sensitivity of a significance test

purpose of a power analysis

determine whether sensitivity level (statistical power) needs to be increased

Power

probability of not making a type 2 error

Odds Ratio

is a way of comparing the odds of an event happening in one group to the odds of it happening in another group.

Relative Risk

actual risk of an event happening in one group or another

significance study

effect size x size of study