1.4 - 2.1 - 2.2

1.4 - Variability and Bias

• Sampling error (Sampling Variability): sample to sample differences/variations

• Sampling Bias: certain outcomes are favored - fair representation = unbiased

Sources of Possible Bias

• Voluntary response bias: leave it up to individual to choose to respond; usually responders have strong opinions abt topic

• Social responsibility bias: dishonest responses if ppl fear judgement

• Leading question bias: confusing/leading questions influencing given answers

• Undercoverage: subpop. is inadequately represented or not sampled

• Nonresponse: ppl asked to participate but refuse; most of sample doesn’t respond

• Self-interest

Ch. 2 - Graphical Summaries of Data

• Frequency table: summary of counts for each category

• Relative Frequency.: freq. divided by sum of all freqs. for a category (

  • rel freq decimals must equal 1/.999/.1.0001 ; rel freq %s equal 100

• Pareto Chart: type of bar chart - categories given w/ bars in descending order

• Side by side chart: 2 bars tg in same category- type of bar chart

2b. Freq. Distributions and Histograms

• Class (bin) width: width b/w consecutive lower limits (ie. 20-10=10); grouping of the data

• Lower class limits: 10, 20, 30, 40

• Upper class limits: 19, 29, 39, 49

• Data: 11, 12, 18, 22, 22, 23, 26, 27, 42, 45, 46, 47

Histograms Shapes

• skewed left = negative skewed

• skewed right = positive skewed

• symmetric = left/right halves of histo. approx. equal

Modality

• Unimodal - one major peak

• Bimodal - two major peaks

• Uniform - no significant peaks