stats vocab

valid

dosen’t equal correct, and is based on the process that is used to make the inference and it needs to be reasonable and transparent — means using math based process to make it

valid inferences

using the information from a small group (a sample) to say something about the whole group (the population)

sample

The collection of individuals from whom information is collected

population

The collection of individuals from whom the researcher would collect information from of was possible and reasonable (the whole group of individuals. the researcher is interested in)

initial

very general starting point for the question

research question

the big overall question the study is supposed to answer

population of interest

the whole group of individuals about whom the researcher wants to answer the research question

measurement

the way info is collected from each individual in the group

statistical measure

the way total data is summarized using a number

sampling methodology

How the sample is chosen- who are the individuals from whom info will be collected chosen

statistical methodology

How the results from this sample (statistical measure) are used to honestly and accurately say something about the population, (specifically the statistical measure for the population)

initial research question

The very early version of the overall question the researchers want an answer to

(going to be very general)

(Measurement) Mode

How dose the researcher interact with each individual (ex: online, face to face, ect)

(Measurement) Form

What dose it look like?
ex: list of questions, surveys, etc

Measures of frequency

intended to describe/ summarize how often a value appears in the data set

Measures of frequency- frequency

(count) How many times the value appears

Measures of frequency- percent

number in the part/ number in the whole (100%) ex 7 divided by 33 × 100 gives you the percent you need

Measures of frequency- population (relative frequency)

the decimal version of the % so: number in the part/ number in the whole = population

measures of center

intended to describe/ summarize what kind of values are in the data set (ex-large tables and small values)

measures of center- mode

the value that appears most often

measures of center- mean/ average

add them up and divide by how many there are

measures of center- median

the middle number in the data set after it is put in order (sorted)

measures of spread

intended to describe or summarize how spread out the value in the data set are

measures of spread- range

how much space the whole data set takes up

measures of spread- interquartile range

how much space the middle half of the data set takes up

measures of spread- standard deviation

how far on average the data points are from the middle of the data set

measures of spread- variance

Instead of distances, the variance looks at squared distances

measures of association

intended to describe or summarize the relationship between 2 values from each individual (ex-, height vs weight and age vs height for people)

measures of association- correlation coefficient (Pearsons R)

Codes the direction of any association (larger values with larger values or large with small) and the strength of association (how often dose the association hold?)

measures of association- proportion of variability

same idea as the correlation coefficient, just squared

Variable

is a trait or characteristic of individuals that can be different for different individuals (i.e. that can vary from individual to individual)

Qualitative

if the possible values represent categories. ex-preferred pronouns (not considered discrete or continuous)

Quantitative

if the possible values are numbers that represent a measurement or a count. ex- age in days

Continuous quantitative variable

there is no “next biggest value” ex- distance from home

Discrete quantitative variable

there is a “next biggest value” ex- shoe size

Parameter

the value of a statistical measure for the population

Statistic

the value of of statistical measure from a sample

Probability

a way of talking about (quantifying) something that has not yet happened

Law of Large Numbers

suppose the probability of a particular result of an “experiment” is P… when the experiment i repeated over and over the proportion of the times the particular result has happened so far will get closer and closer to P in the long run

experimental approach

if an experiment has been repeated a large number of times and the proportion of times a result has happened is P… then the probability of the results happening is approximately P

The random pick/ relative frequency approach

Suppose in a population the proportion of individuals that are X is P. If we picked one individual at random from the population the probability of getting one that is X is P

sampling methodology

a method for selecting the individuals to collect info/data from

statistical methodology

a math-based method for making an inference about the parameters

(errors of sampling) sampling error

if a sample is used in place of the population, the result will be wrong

(errors of sampling) sampling bias

from when the way the sample was chosen is biased

measurement

go to each individual in the sample and gather information/ data

confidence interval

use a statistic to estimate the parameter

(confidence interval) Point estimation

a single value that is probably close to the parameter (i.e. the statistic)

(confidence interval) Margin of error

a distance that represents how close “close” is

(confidence interval) Confidence level

How confident we are that the statistic is “close” to the parameter

hypothesis test

compare the parameter to a value or another parameter using a statistic or statisitics

(hypothesis test) Hypothesis

the “direction” that we think comparison goes (greater than, less than, equal to, ect)

(hypothesis test) P-Value

The probability of getting a result like the actual statistic/s assuming the comparison is actually an equals

Correlation

use a sample to determine if there is a link (an association) between 2 variables

(correlation) Correlation Coefficient

This means 2 things :

1) the “direction” of the association, positive or negative

positive: means larger values of one variable usually go with larger values of another variable

negative: larger values of one variable tend to go with smaller values of another variable

2) the “strength” of association, how often association actually holds