Module 8 - Quantitative Analysis

• Nominal

• Ordinal

What 2 levels of data use nonparametric statistical tests?

• Interval

• Ratio

What 2 levels of measurement use parametric statistical tests?

Nominal Level Data

Used when assigning numbers to different categories s

There are different groups

What does nominal data allow us to determine?

• Severity ratings

• Difficulty ratings

• Satisfaction

What are some things that ordinal data is good for?

Order

What needs to be maintained in ordinal data?

Inequality

What is there a property of in ordinal data?

• IQ scores

• Standard scores

• Language scores

What are some examples of interval level data?

• Duration

• Frequency

• Intensity

What are some examples of ratio data?

Equal

Is there an equal or unequal distance to the mean in interval level data?

Magnitude

In interval level data, you can describe THIS by adding or subtracting numbers.

Ratio

What type / level of data is the most robust with the most qualities that allow you to do better analyses?

• Adding

• Subtracting

• Multiplying

• Diving

How do you describe magnitude using ratio level data?

Normally distributed data

Parametric tests are ideal for THIS.

Distribution Free Tests

Nonparametric tests are ideal for THIS

Shape of the distribution

What partially drives the choice of statistical methods?

Parametric

Between parametric and nonparametric, what is seen as more powerful?

• Types of data

• Number of participants

• Normal distribution

• Homogeneity of Variance

What are some assumptions when looking at statistical tests?

Normal distribution theory

Scores shouldn’t be all over the place and they should be teased out to look at normal distribution

Type of Data

• What is the level of measurement

• Nonparametric tests are used for nominal and ordinal

• Parametric tests are used for interval and ratio

Number of Participants

• Nonparametric tests are used when the sample is small (<30)

• Parametric tests are used when the sample is large enough (>30)

• If the sample size is not big enough, you may not be able to detect significant differences between the groups

<30

What is considered a small sample?

>30

What is considered a large sample?

Small

Are nonparametric tests used with a large or small sample?

Large

Are parametric tests used with a large or small sample?

Normally distributed

Are parametric tests normally or not normally distributed?

Not normally distributed

Are nonparametric tests normally or not normally distributed?

Homogeneity of Variance

Implies there is a similarity between groups in terms of variation

Less variability

If a group has a mean of 50 and a standard deviation of 5, do they have more or less variability?

More variability

If a group has a mean of 70 and a standard deviation of 20, do they have more or less variability?

Assumptions

What helps the researcher know which statistical test to pick?

• Level of measurement of the DV

• Experimental design

• Other assumptions being met

What does the selection of analytical procedures depend?

Homogenous

What should variances be?

Unrelated samples

2 groups aren’t made up of the same subjects

Between-subject designs

What is tests for unrelated samples also called?

Post-Hoc analyses

Tells us where the statistical significance is in the study

Main effect

Measurement of statistical significance somewhere in a complex study and is the first analysis that is done looking for statistical significance in your complex study

Probability

You compare the calculated t value to the critical value to determine if there is a statistical significance

Probability Level

Reported so the researcher can determine if the differences among the groups are due to the experiment or due to chance