


Draw a sample from a larger population to represent the whole population. Put names in a hat. Make sure every member/element has equal chance of being selected. 


1/2 of random sample goes to treatment group and the other 1/2 goes to the control group. 


entire collection of events (scores, incomes, speeds, etc.) that you’re studying 


small number from larger population. Allows us to infer something about characteristics of population. 


1st aspect of randomness. Does the sample reflect the population? Ex. Small town from NE wouldn’t rep US hispanic culture well. 


2nd aspect of randomness. After the subjects have been selected, subjects then have to be randomly assigned to treatment or control. 


Are the results the result of the differences in the way we treated our groups (hope so) and not a result of WHO we placed in each group. 


property of an object/event that can take on different values (hair color, selfconfidence, gender, personal control, treatment groups) 


Researchers decide what these will be (group memberships like gender groups or teaching style, etc.) 


Researchers have NO control over these. (resulting selfesteem scores, personal control, etc.) 


limited number of values (gender, high school grades) 


any value between lowest and highest points on a scale (age, selfesteem) 


aka. Measurement Data.Numerical data (weights, test scores) other “how much” tests 


aka. Quantitative Data. Numerical data (weights, test scores) other “how much” tests 


aka. Qualitative/Frequency Data. (no numbers) 


aka. Qualitative/Categorical Data. (no numbers) 


aka. Categorical/Frequency Data. (no numbers) 


Describing a set of data (means, graphs, extreme scores, oddly shaped distributions) 

Exploratory Data Analysis (EDA) 

Joh Tukey showed necessity of paying close attention to examining data in close detail before invoking more tech involved procedures 


do after descriptive statistics, after we have a basic understanding of the numbers. 


a measure that refers to an entire population (average selfesteem score) 


same measure as a parameter, calculated from sample of data we have collected. 


labels categorical data. (gender, political parties) 


simplest true scale. Orders people, objects or events along a continuum (ranks in military) 


allows us to speak of differences between scale points (same different between 1015 degree C as there is between 1520 degree C) 


has a true 0 point (true absence). Allows us to speak of ratios/fractions. (length, volume, time) 
