Ch 1 pt 5 - Graphs & Distribution Shapes Flashcards
Bar graph
For data measured at nominal or ordinal level, or discrete data.
Bras in a bar graph don’t touch
Histogram
Used when continuous data is measured on interval or ratio scales.
Looks like a bar graph, but the bars touch.
The height of the bear represents frequency measure.
Controversies & limitations of graphs
- Unequal interval sizes
- Exaggerating proportions
- Outliers: extreme scores that are more extreme than most of the scores in the data set.
Peak
represents the most frequently occurring value in the data set
Shapes of Freq. Distributions: Unimodal
One clear peak
Shapes of Freq. Distributions: Bimodal
Two peaks
Shapes of Freq. Distributions: Uniform/ rectangular
Doesn’t have any clear peaks, and all values are about the same.
Shapes of Freq. Distributions: Multimodal
More than two peaks
Symmetrical distributions
One in which it appears you could fold it in half on its vertical axis, and the two sides would pretty much match.
Skewed distributions
Where scores pile up on one side of the distribution.
We describe the skew in terms of the direction of the tail.
Positive skew
The tail is in the positive direction
Most scores pile up on the negative end
Lots of low scores, fewer high scores
Negative skew
The tail is in the negative direction
Most scores pile up on the positive end
Lots of high scores, fewer low scores
Kurtosis
Measures the “fatness” of the tails of a distribution and how peaked or flat the distribution is.
Platykurtic
negative kurtosis
flatter peak, thinner tails
Mesokurtic
0 kurtosis (e.g. normal distribution)
