Test for Normality and Scales of Measurement

ANALYTICALLY, GRAPHICALLY

In our previous lesson we discussed how to test for the normality of a data which can be done __________ or _________.

Quantile-Quantile plot (QQ)

Graphically we can use what we call a ______ plot.

Quantile-Quantile plot (QQ)

Method where you will be plotting the data as if it was a scatter plot and testing the data by the use of a diagonal line test.

the more that it is probably normally distributed

The more that the plot is closer to the diagonal line test, ____________.

more

The _____ that the plot is closer to the diagonal line test, the more that it is probably normally distributed.

diagonal line test

The more that the plot is closer to the ________, the more that it is probably normally distributed.

more

The more that the plot is closer to the diagonal line test, the _____ that it is probably normally distributed.

Analytical method

What is the more efficient way to test for normality?

Analytical method

This can be done by the use of the Shapiro Wilk Test and Kolmogorov Smirnov.

Shapiro Wilk Test, Kolmogorov Smirnov

What tests do you need to do with the Analytical method?

Levels of measurement

There are two kinds of this: categorical and numerical.

Nominal

This level of measurement is the MOST BASIC level of measurement.

nominal

In this data, you will not literally be measuring anything as if this data is identified by classification, groups or characteristics of the data.

nominal

What type of data is the following:

- Genders of respondents

- Hair Colors.

- Favorite Fast Food Chain, etc.

Ordinal

This data can still be classified by their characteristics but in a more meaningful order about the nature of the data and has a differentiated value per each rank.

ordinal

What type of daya are the following:

- Levels of Satisfaction(Very Satisfied, Satisfied, Not Satisfied)

- Levels of Income(Low income, Medium Income, High Income)

- Happiness on a scale of 1-10 (this is what's known as a Likert scale)

numerical data

This data is used for quantitative research wherein this data is naturally measured in numbers.

Interval

This type of level of measurement also has an order like ordinal data wherein the spaces between points are equal.

interval

What type of data is the following:

Grades (88-90, 80-82)

interval

In this type of data there is no such thing as true zero where all numbers in this data have a meaning including zero. Which makes this data more sophisticated than the other types of data.

ratio

The Jack of all traits in the level of measurements wherein all of the characteristics that the other three measurements are included in the Ratio level of measurement.

ratio

There is a true zero in this type of measurement where the zero means null or nothing which serves the true meaning of a zero.

Chi^2

What test should you use if:

V1: Nominal/Ordinal

V2: Nominal/Ordinal

T-Test

What test should you use if:

V1: Nominal/Ordinal

V2: Interval/Ratio

Parametric

Mann Whitney U-Test

What test should you use if:

V1: Nominal/Ordinal

V2: Interval/Ratio

Non-Parametric

Pearson-r

What test should you use if:

V1: Interval/Ratio

V2: Interval/Ratio

Parametric

Spearman rho

What test should you use if:

V1: Interval/Ratio

V2: Interval/Ratio

Non-Parametric