7 Oct 2016

School

Department

Course

Professor

STAT*2060: Statistics for Business Decisions

Week 1 Lectures

1 What is Statistics?

1: A characteristic of a sample (we will come back to this shortly!)

2: The science of collecting, analyzing, interpreting, presenting, and making conclusions

from data.

Statistics is a science that allows us to answer questions such as:

•Does there exist a relationship between the amount of money spent on an advertising campaign,

and the resulting sales?

•Does one chemotherapy drug result in higher proportions of colon cancer survivors than another

chemotherapy drug?

•Can a grocery store chain use population demographics (population size, household income, house-

hold demographics (kids, no kids, pets, etc.)) to predict average monthly revenue?

1

To answer these (and other!) questions, we must:

2 Data Collection: Terminology and Techniques

Example: Why the hate-on for Pokemon Go? It’s making people healthy

http://www.cbc.ca/news/business/pokemon-go-exercise-1.3727123, Aug. 21, 2016.

An article on CBC.ca discusses the Pokemon Go craze, and brings into question the social media

backlash against Pokemon hunters when there appears to be a positive health beneﬁt associated with

playing the game. In particular, the article makes the statement “Sixty per cent of American millennials

surveyed in a new Manulife poll said playing the game has upped their activity levels”, and links to the

following info-graphic:

2

Let’s use this example to deﬁne some important terms:

Population:

Sample:

(Experimental) Unit:

Parameter:

Statistic:

In statistics, we can use information from our sample data, and our sample statistics, to make state-

ments about the corresponding population parameters. This is a branch of statistics known as inferential

statistics, which we will come back to in a few weeks.

The question then arises: “How do we obtain a sample, or collect sample data?”

2.1 Sampling Methods

There are various sampling methods that can be used to obtain a sample from a population. Which

method we choose can depend on a variety of factors (resources, what we are trying to sample, etc.).

However, in general the preferred sampling methods are those that incorporate some degree of ran-

domization, as this helps to limit any potential sampling bias that may exist.

Simple Random Sample (SRS):

3