Home > Preview
The flashcards below were created by user
Ehl
on FreezingBlue Flashcards.

Standard Variation
Variance
Measure of spread
 Measure of spread
 (Standard Deviation)^{2}

What are the measures of center?
Mean, Median, Mode

Describing the data: Spread (describe)
Is the data spread out or clumped?

Describing the data: Center (what is it?)
The middle value

How do you find outliers?
 Inter Quartile Range (IQR)
 It is found by Q_{3}Q_{1}
 Next take the IQR and multiply it by 0.5
 Now add that number to Q_{3} and subtract it from Q_{1}

When should outliers by mentioned?
Always say how many outliers there are or none if there aren't any.

When describing data, how is symmetry mentioned?
Symmetrical or not. If no, scewed left or right depending on the data.

What are the different types of modes when describing data?
Unimodal, bimodal, multimodal, uniform

Qualitative Data (+example)
data that is sorted into categories. Pie/bar graphs are used. Ex: Gender

Quantitative data (+example)
uses numerical values. Histograms, stem and leaf, dot plot, frequency table. Ex: numbers

Population related to sampling
 The entire group of individuals (who) that we want to gather information on
 We define the population

Sample in terms of sampling
part of the population we can gather info on. There are many different ways and the cost/time are efficient

Census
Gather info on the entire population. It is a more accurate method but can be expensive and time consuming. The population may change making it hard to know when finished.

Types of Sampling: Simple Random Sampling or SRS
SRS is conducted by randomly picking individuals using number generators, flipping a coin, rolling a die, etc.

Types of Sampling: Stratified Random Sampling.
The population is put into strata (subgroup) and are randomly sampled. Examples of strata are age group, political party, gender, etc. Works in tandem w/SRS

Types of Sampling: Representative Sampling.
 Strata are defined and a proportional
 sample is taken from each strata. It is a sub type of Stratified Random Sampling and works in tandem w/SRS. It is the best but hardest type to perform

Types of Sampling: Cluster Sampling
 The population is put into subgroups (not
 strata). The clusters (subgroups) are chosen at random, and the whole cluster is sampled. The Subgroups should be representative It works with SRS but clusters have to be representative.

Types of Sampling: Systematic Sampling
A system is set up to select individuals instead of using random methods. Ex: Every third individual is selected. Used in product production but may be biased.

Natural Variation
Variation that happens in the sample because there is variation in the population. Using the appropriate technique can limit part of the population being over represented.

Explainable Variation
Variation due to different stratas. Ex: gender, status, grade, where you live, etc.

Sampling Variation (error)
Exists purely because you are taking a sample and therefore want to get 100% accurate representation of the population.

Biased Sampling Error
True error and can be due to many things such as: sample size, method of sampling, and survey questions.

Confidence Interval?
This interval gives an estimated range in which we believe the actual (census) numbers would fall into and is usually a %. Margin of error.

Confidence Level?
The degree in which you have confidence that the numbers you obtained are in that interval and is expressed by saying what others should get if they were to do the same sample

Confidence Level and Interval are related to _______
The confidence level is usually __% and __% (3 Standard Deviations away from the mean)
The higher the ___, the larger the ___
 Standard Deviation
 95%; 99%
 Level;interval

Correlation and formula for r?
It describes the relationship of different data is described as strong/weak positive/negative.

Regression Line and its equation
It is used to extrapolate and interpolate data.
y=mx+b

response Bias
When the survey design influences results; the most common type of response bias comes from wording of the questions.

What makes a good question?
 Only asks one question.
 can accommodate all possible answers
 No ambiguity
 Variety of Answers
 Doesn't assume too much

non response bias
Bias when a large group doesn't respond.

voluntary response bias
A type of non response bias where particular participants chosen may only be ones who care.

Area Principle in graphs
When some visual representations are not proportional/same dimensions.

