Lecture Week 4

Question 1

Dichotomous Variables

Answer 1

• A categorical value
• variable has 2 groups

eg: a treatment group and a control group

Question 2

Other names for a Scale Variable

Answer 2

• SPSS combines Interval and Ratio variables into Scale
• Continuous variable
• Metric Variable

Question 3

Other names for Categorical Variables

Answer 3

• Grouping Variables
• Nominal Variables
• Non-Metric Variables

Question 4

Polytomous Variable

Answer 4

• Poly means many
• Means more than two groups or many groups

Question 5

Three types of t-tests

Answer 5

1. Single-Sample t-tests (sometimes called one-way or simple t-tests)
2. Independent Samples t-tests (sometimes called Independent Groups t-tests)
3. Paired or Matched or Repeated samples t-tests

Question 6

What is a t-test?

Answer 6

• Asks if there is a significant difference between two means
• Used to test questions around mean differences

Question 7

William Gosset

Answer 7

• Invented the t-test because he wanted to improve productivity of time in motion in the brewery
• Part of a group of mathematicians in latter 19th/early 20th centuries including Pearson, Popper, Ronald Fisher, Tukey
• Considered golden era of statistical thinking
• Worked at the Guiness Brewery in Dublin
• Wrote under the pseudonym of Student

Question 8

Before you run a statistical test …

Answer 8

• Run the Descriptive Results to find the means
  
  • Analyse/Compare Means
• Run a graph
  
  • Graphs/Legacy Dialogues/Bars Represent/Other Statistics (eg: means):_Variable/Compare Axis: _Variable​

Question 9

Three statistics to focus on in t-tests

Answer 9

• t Score - The actual t value
• Degrees of Freedom - df -
• Significance Level - this is the p value - tells us directly whether there is a statistically significant difference.

Question 10

What are Degrees of Freedom?

Answer 10

• (no. of subjects group 1 -1) + (no. subjects group 2)

or

• Total Subjects - 2
• Refers to the maximum number of logically independent values that have the freedom to vary in the data sample.
• Commonly discussed in relation to hypothesis testing such as a Chi-Square.
• Calculating df is key to understanding the importance of a Chi-Square statistic and the validity of the null hypothesis.