Stats- Monday 1/22/18
Chapter 1- Defining and collecting data
DCOVA- different steps of the process of stats.
D-Defining
C- Collecting data
Categorical (qualitative) variables
Labels or categorical names are used to take categories as their values
Two kinds of variables-concept that makes sense to us
1. Categorical Qualitative- used to categorize values, referring to words, names, major, sex
2. Numerical/ quantitative- anything to do with numbers, measuring,
a. Discrete values- derives from a counting process, number of job offers
b. Continuous- measuring process, metaphor to use, “always think about
measurements, for example a ruler, they typically have intervals, month to
month, inch to inch, ounce to ounce, Mega byte, time”
Where can we collect data?
1. Primary source- statistical studies, experimental or observational
2. Secondary- gathering from something NOT first hand, employee records, survey
responses
Research focuses on population
– Target populations include items or individuals that you want to reach a conclusion
Variable
Structured vs. Unstructured data
Electronic editions of data
Wednesday January 24th
Defining and collecting Data
Distinguishing between the types of sampling and methods and to understand surveys and
ethics
Nonprobability and Sampling Methods
Sample- small group of a whole
Random or NON Random??
Non-probability
Convenience sampling- seleting easy items that are convenient to the researcher/ Draw backs
are that youre not gathering entire population. Once you know your target population and
what questions you want to ask. Easy targets to get, for a professor it would be their students.
Judgement sampling- Nonrandom. You collect data from pre-selected experts in the subject
matter but this sample cannot be generalized. Gather a group of doctors and patients and find
out abou the topic of research because they are experts.
Probability sampling methods– focuses on random chance
Simple random sample- every item from a frame has the same chance of selection as every
other item.
Two approaches with randomnesss
Theres 100 houses but youre only going to asses 20 of them.
1. Sampling without replacement: we include properties for only one time to be in our
