2.1 Errors in Sampling

Key Terms

statistics - the branch of mathematics that describes the gathering, organizations, analysis and presentation of numerical information

population - entire group of individuals

census - collection of info from a population

sample - a subset of the population

survey - the collection of info from a sample

raw data - unprocessed information collected by study

variable - the quantity being measured

numerical variable - a variable that has a numerical and can be counted or measured

categorical variable - a variable that is described by some non-numerical characteristic that is divided into categories

discrete variable - a variable that can have only certain separate values within a given range (data that can described using whole numbers or integers)

continuous variable - a variable that can have any value within a given range (data that are only measurable with real numbers) ex. height, weight, time, temperature

Bias

The best way to same is to obtain a random and representative sample group. When factors cause the sample to not be representative we introduce bias.

A statistical bias is any factor that systemically favors certain outcomes in a survey or experiment

Population values therefore not be accurately reflected in the survey/experiment results

Types of Error

Sampling Error

  • occurs when the sampling frame doesn’t reflect the charcteristics of the population

  • occurs randomly, even when samples are selected to avoid bias

Measurement Error

  • occurs when the data-collection method consistently under- or over-estimates a characteristic of the pop.

  • consistent measurement errors skew the results of a survey

Coverage Error

  • occurs when the same is not sufficiently large and unbiased

Non-Response and Response Errors

  • occurs when particular groups are under-represented in a survey because they choose not to particiate, or otherwise Amy feel uncomfortable answering truthfully

Leading and Loaded Questions

  • questions asked in a survey intended to prompt specific answers

  • questions will include or imply the desired answer

  • contain wording or information intended to influence the respondents’ answers

  • tends to contain strong or emotive language

Sampling Methods

1. Simple Random Sample

Everyone in the pop. has an equal chance of being selected. ex. drawing names from a hat randomly for survey.

2. Systematic Sample:

Selecting members at regular intervals

Interval = pop. size/sample size

3. Stratified Sample and Quota Sample:

Divide the population into groups (called ‘strata’) and select a number of individuals from each group proportional to their size in the population

4. Convenience Sample:

Members of pop. are selected as a sample simply because they are easily accessible