PSYC220 - spring2023 exam 1

Quantitative psychology

application of statistical methods to psychology

Psychometrics

application of statistical methods to psychological measurement; construction of measurement instruments

Defined by the research question

population

Often not possible to measure everyone in the population

population

Subset of the units of a population

sample

A value describing a population

parameter

A value describing a sample

statistic

Sample mean in psychology is "M"

statistic

Goal: Improve understanding of some phenomenon or solve practical problem

research project

descriptive statistics

summarize and displays the data
Examples: frequency tables, histograms, bar graphs, box plots

inferential statistics

explanation or prediction
Examples: hypothesis testing (example = t-test), regression, ANOVA

Descriptice statistics (raw scores)

Look for patterns
Summariza data
Present information in convenient form

inferential statistics

Make estimates, decisions, predictions, generalizaitons about a larger set of data

correlational method

is there a relationship between a pair of variables?

experimental method

demonstrate a cause-and-effect relationship among variables

independent variable

variable that is manipulated by the researcher. like pictures of faces, houses, chairs

quasi independent (non-experimental studies)

variable that is not manipulated by the researcher, but is used to create different groups

Dependent variable

blood oxygen level dependent response

dependent variable

observed

Statistics

the best estimate of sample

The larger the sample the more accurate the _______________

parameter

sample

statistic

population

parameter

population

defined by research question

correlation studies

observing number of triggers and vividness of mental imagery

experimental studies

variable that we manipulate (independent variable)

dependent variables

how quickly someone can read the word and react to it

independent variable

type of word someone is exposed to

Descriptive statistics

sample

inferential statistics

population, when you say something about the population

nominal scale

handedness: left-handed, right-handed, ambidextrous

ordinal scale

small, large, medium

nominal scale

qualitative

ordinal scale

qualitative

Interval scale

temperature, sea level, time grades, quantitative

Ratio

reaction time, absolute 0 point, height, weight, income, age

ratio

quantitative

mean

add all the values together and divide by how many there are

Frequency Distribution

a tabulation of the number of observations located in each category on a scale of measurement

Discrete and qualitative Data

Frequency distribution table
Bar graph

Continuous Data

Frequency distribution table
Histogram

class

one of the categories (qualitative data) - type

frequency

number of observation in a class - number

relative frequency

class frequency divided by total number of observations - proportion

continuous

doesn't matter how close the values are, it always goes to infinity. age, can be measured on ratio scale

continuous data

there are infinite number of possible values that exist between any two observed values

outlier

score that is extreme compared to other scores in the data set (to the point of being suspect)

question to ask about outlier

why did the outlier occur

why did the outlier occur

"Type" (unrealistic observation)
Unique observation (true measurement)

How to handle outliers?

Correct (only for "typos")
Keep
Remove (implications for generalization that are made)

it is impractical to construct a frequency distribution with too many values

rationale

How many groups to select?

Too few vs. too many

What interval size to pick?

Highest score - lowest score, divide by the desired number of groups, round

What should be the beginning value of the lowest interval?

Select a value that is smaller than the lowest score
Number that is evenly divisible by the interval size

histogram

condenses the values by grouping similar values in the same class in the graph
represents frequency or % by area

Frequency distribution

tables and graphs

types of graphs in frequency distribution

bar
histogram

Most representative value / central tend

mean
median
mode

Central tendency

Tendency of a data to cluster about certain numerical values

median

the position in the distribution that divides the ordered set of scores in two equal groups

computation

arrange measurements from smallest to largest

mode

measurement that occurs most frequently in the data set

Median

value that is exactly in the middle

variablity

how the scores are different (spread out or cluster together) from one another

range

difference between largest and smallest value

interquartile range

range covered by the middle 50% of the distribution

semi-interquartile range

half of the interquartile range

st. dev.

approximates the average distance from the mean

variance

sum of squared deviations / number of scores - 1

z-score

Z = (x-mean)/st. Dev.

Population st. dev

sq. root of variance

sum of squares (ss)

measure that quantifies the variability or dispersion in a set of data. The sum of squares is calculated by summing the squared deviations of each data point from the mean. It is commonly used in various statistical analyses, such as calculating variance and standard deviation

Describe the scores in a sample that has a standard deviation of zero.

it means that all the scores in the sample are identical

What information is provided by the sign (+/-) of a z-score?

the direction of the score relative to the mean. A positive z-score indicates that the score is above the mean, while a negative z-score indicates that the score is below the mean

What information is provided by the numerical value of the z-score?

the distance between the score and the mean in terms of standard deviations

Mean square distance

variance

If the score is larger than st. dev. The z-score is going to be ___________

bigger

sampling error

the difference between the sample result and the actual population result

Explain why this phenomenon (sampling error) creates a problem to be addressed by inferential statistics

Because sampling error can lead to inaccurate conclusion about the entire population, which inferential statistics aims to address