College Statistics Module 1

variable

something that holds values

data

value the variables take on

qualitative data

qualities, non-numerical categories

quantitative data

represents counts or measurements

population

all subjects in a study

sample

subjects selected from the population

population parameters

numbers or observations that describe something about the population

sample statistics

numbers or observations that describe something about the sample

representative sample

relevant characteristics of sample members that match those of the population

simple random sample

selecting sample elements randomly; every element in the population has an equal chance of being selected for the sample

systematic sampling

following a system in selecting the sample such as selecting every Kth person in the line or list

stratified sample

selecting number groups randomly, such as males or females, age divisions, religion preferences, then selecting a subset of each group randomly (approximately equal ratios)

cluster sampling

selecting number of groups randomly, then take all elements of the selected group/s to be the sample (such as all students in a certain class period or everyone on the 5th floor)

convenience sampling

selecting a group of subjects that is convenient to sample

observational studies

observe characteristics of sample members without attempting to influence or modify it, then drawing conclusions based on these observations

experimental studies

manipulating the independent variable to determine the effect on the dependent variable; researchers divide the sample into two or more groups and treat every group differently

treatment group

receives experimental treatment

control group

receives a placebo

blinding techniques - single blinding

participants don’t know the group they belong to, but researchers do

binding techniques - double blinding

neither participants nor researchers know who belongs to which group

nominal

qualitative categories with no order or rank such as a city of birth, gender, ethnicity, etc

ordinal

qualitative categories have order or rank (intervals between rankings might vary) such as top 5 olympic medalists, likert-type questions (very dissatisfied to very satisfied), etc

interval

quantitative data can be categorized, ranked, evenly spaced, has no true zero such as Celsius or Fahrenheit

ratio

quantitative data, can be categorized, ranked, evenly spaced, has as natural zero (an absence of the variable interest), such as height, age, weight, temperature in Kelvin (no negatives)

class limit

smallest and largest values that can belong to each class in a frequency distribution (lower class limit & upper class limit = lower class limit + class width - 1)

class width

(highest value - lowest value) / number of classes

*round answers up to the next whole number

class boundaries

lower limit - 0.5; upper limit + 0.5

midpoint

(lower limit + upper limit) / 2

frequency of a class

number of data values in that class

cumulative frequency

frequency of the class + ALL previous frequencies

relative frequency

frequency of the class / total number of values

relative cumulative frequency

cumulative frequency of the class / total number of values

total number of values

sum of frequencies