Week 12 – Descriptive statistics Lectures

Flashcards on descriptive statistics based on lecture notes.

Descriptive (univariate) statistics

Summarizes the data on one variable at a time; for example, the number and percent of students by gender.

Bivariate Statistics (crosstabulation)

Summarizes the relationship between two variables without any hypothesis testing; for example, the number and percent of students by gender identity and race/ethnicity.

Descriptive (univariate) statistics

Summarizes the data on one variable at a time; for example, the number and percent of students by gender.

Bivariate Statistics (crosstabulation)

Summarizes the relationship between two variables without any hypothesis testing; for example, the number and percent of students by gender identity and race/ethnicity.

Frequency (n)

Measures of distribution for nominal and ordinal variables, representing the number of participants with a certain characteristic.

Proportion

measure of distribution. The frequency of participants with a certain characteristic (n) divided by the total number of participants (N) surveyed, ranging from 0 to 1 and typically reported as a decimal.

Percent (%)

measure of distribution.The proportion times 100, representing the number of people per 100 with a particular characteristic; calculated out of the people who participated in the study. report percent> valid percent bc its more accurate

Valid Percent

Percent calculated out of the people who were in the study and answered the particular question, ignoring missing values.

Mode

measures of central tendency. The most frequently occurring values; rarely reported.

Mean (M, or x-bar)

measure of central tendency. The average score across all participants, sensitive to extreme scores, reported when a variable is normally distributed or for individual/composite scored variables.

Median (med.)

measure of central tendency. The middle score or 50th percentile (Q2) across all participants in a skewed distribution, or with ordinal Likert-type scales.

Standard Deviation (s, SD)

measure of dispersion or variability. The extent to which participants' scores differ from the mean; reported with the mean and sensitive to extreme scores.

Quartiles

measure of dispersion or variability. (1st quartile or Q1) and 75th (3rd quartile or Q3) percentiles, reported with the median (2nd quartile or Q2)

Range

measure of dispersion or variability. The difference between the highest and lowest score; we report the observed lower and upper limits not the actual range.

Interquartile Range (IQR)

measure of dispersion or variability The difference between the 75th and 25th percentiles; range provided for the median; IQR = Q3 - Q1.

Descriptive statistics to run and report for nominal and ordinal data

Frequency and percent.

Descriptive statistics to run and report for normally distributed continuous data (interval and ratio)

Measures of central tendency and dispersion (variability) – pick either mean and SD depending on distribution (pick SD or mean when there is normal distribution) or Median and quartiles (report when there is skewed variable)

Visuals for nominal and ordinal data

Bar graph or pie chart

Visuals for continuous data

Histograms (positive skew to the right, normal bell-shaped, negative skew to the left)

Frequency Distribution

A table that provides information about frequencies for each score or characteristic in a variable.

How do you report descriptive statistics?

In methods and results sections there is no interpretation just state the facts.