unit 10 Non-Parametric Statistics

A set of flashcards covering vocabulary and definitions pertinent to Non-Parametric Statistics as discussed in the lecture notes.

Non-Parametric Procedures

Statistical methods used for nominal (categorical) and ordinal (ranked) data.

Assumptions for Non-Parametric Data

Data should be interval or ratio; expected to have a normal/near normal distribution.

Robust Procedures

Parametric procedures considered robust are designed to tolerate certain degrees of violation of normality and homogeneity.

Chi-Square Test

An inferential procedure used with nominal data to examine relationships between categories.

One Way Chi-Square

A chi-square test with one independent variable, requiring at least two mutually exclusive categories.

Expected Frequency (fE)

The frequency that is predicted based on the null hypothesis for each category.

Goodness of Fit Test

A test comparing observed sample data and the null hypothesis against how well they fit.

Degrees of Freedom (v)

A calculation used in chi-square tests to determine the critical value needed for significance.

Two Way Chi-Square

A chi-square test examining frequencies across two factors (2 iv)

while maintaining the same assumptions as one way.

Test of Independence

Calculates if the categories of two factors are independent of each other.

Mann-Whitney U Test

A non-parametric test used to determine differences between two independent samples of ranks.

U(obtained)

The calculated value from the Mann-Whitney U test used to determine significance.

• it is a 2 tailed test

• compare the obtained values and the critical values found in the table of the mann whitney

• if u is less then or equal to the critical u then this is a significant result

• the smaller u the more it is that the null hyppothesis is false and will be rejected

Wilcoxon T Test

A non-parametric test for two related samples (repeated measures) assessing ranked data.

-if T obtained is less than the critical it is significant

Difference Scores

Scores obtained by subtracting one measurement from another in paired samples for ranks.

Kruskal-Wallis H Test

A non-parametric equivalent to the one-way ANOVA for comparing more than two independent groups.

Null Hypothesis

A statement predicting no effect or difference in a statistical test.

Alternative Hypothesis

A statement predicting an effect or difference in a statistical test.

Critical Value

The value that a statistic must exceed in order to reject the null hypothesis.

Rank Assignment

The process of ordering data points to transform them to ranks in non-parametric tests.

Significance Level (alpha)

The threshold used to determine whether to reject the null hypothesis, often set at 0.05.

Type I Error

The error of rejecting a true null hypothesis, also known as a false positive.

Homogeneity of Variance

The assumption that different groups have similar variances.

Skewed Distributions

Distributions that are not symmetric and can impact the mean and statistical results.

Two-tailed Test

A test that checks for the possibility of an effect in two directions, positive and negative.

Outlier Impact

The influence of extreme values that can skew results and statistics.

Observed Frequency (fO)

The actual count of occurrences in each category from the data collected.

Independence in Statistics

A condition where one variable's occurrence does not affect another variable's occurrence.

Ranks for Non-Zero Differences

In the Wilcoxon T test, ranks are assigned only to the differences that are not zero.

Interaction between Variables

The relationship or effect that one variable has on another in a statistical analysis.

N (sample size)

The number of observations or data points in a statistical sample.

Critical U Value

The value used to determine the significance of U(obtained) in Mann-Whitney U tests.

Rank Signs in Wilcoxon

The differentiation of ranks based on whether difference scores are positive or negative.

Comparing Two Independent Samples

Evaluating the differences in ranks or values between two separate groups.