Introductory Statistics for Psychologists

Lesson 3: Gathering and organizing data:

Tables, Charts, Shapes of Distributions

Michal Kanat-Maymon, PhD

 Frequency Tables

 A data organization method that displays the values of a

qualitative or quantitative-discrete variable with a relatively
small number of values.
Frequency tables – Nominal

Position of 150 company employees:

X f(x) P(x)
Position Frequency Relative
Frequency ( Percent %)
Administration 30 30/150*100=20%
Sales 60 40%
IT 20 13.33%
HR 10 6.67%
Other 30 20%
Total: N=150 100%
Frequency Tables – Ordinal or Quantitative
Discrete with few values

Number of children in a sample of 200 families:

X f(x) P(x) CF(x) CP(x)

Number of Frequency Relative Cumulative Cumulative
children Frequency ( Percent %) Frequency Percents

0 20 20/200*100=10% 20 10%

1 30 15% 50 25%
2 50 25% 100 50%
3 60 30% 160 80%
4 40 20% 200 100%
Total: N=200 100% N=200 100%
Frequency Table - Terminology

X Variable Each row represents a value

Number of participants in the

f(x) Frequency
category/ value

Relative frequency Percentage of participants in the

P(x)
(percent) category / value

Number of participants up to and

CF(x) Cumulative frequency
including the category / value

Percentage of participants up to and

CP(x) Cumulative percent
including the category / value
 Ages of 294 participants:

35 19 19 26 35 26 18 61 18 32 18 24 18 25 61 18 32 19 26 35 28

28 42 61 18 32 45 16 32 21 28 42 61 18 32 45 16 32 61 18 32 19

16 45 45 42 16 32 21 22 18 30 21 32 37 23 36 32 37 23 43 47 61

41 38 32 37 23 19 35 23 41 34 35 19 26 28 37 19 26 28 19 26 35

23 38 19 26 28 61 54 28 21 35 54 61 18 32 26 69 18 32 61 18 32

28 42 61 18 32 45 16 28 42 61 18 32 45 16 32 37 23 19 26 35 28

32 37 45 41 19 38 23 19 54 41 33 19 26 35 19 26 28 61 18 32 28

19 26 38 21 61 42 28 61 28 42 61 61 18 32 61 18 32 27 48 53 19

61 18 33 35 45 37 32 45 28 42 61 18 32 19 26 35 44 57 23 41 61

28 42 61 18 32 45 19 26 35 54 32 37 23 61 18 32 52 32 28 21 35

38 21 27 19 26 35 61 18 32 16 19 26 28 32 37 23 19 26 35 19 38

42 28 42 61 18 32 45 16 32 32 61 18 32 19 26 28 61 18 32 19 38

37 54 31 54 19 26 35 28 42 61 18 32 45 45 28 42 61 18 23 41 34

26 10 58 25 61 45 28 42 61 18 59 45 18 45 28 42 61 18 28 21 35
Grouped frequency tables –
Many quantitative values

Frequency Table Grouped Frequency Table

X f(x) X f(x)
Number of Frequency grades Frequency
children

0 20
50 - 59 2
1 30
60 – 69 8
2 50
70 – 79 14
3 60
80 – 89 23
4 40
90 - 100 18
Discrete values
(a small range of values) Classes of values
(a large range of values)
Grouped Frequency Table:
Pros and Cons
Pros:

A clear and simple organization of a substantial collection of data.

“bird’s-eye view”

Cons:

Sometimes, it provides a too general picture.

Loss of specific data

(e.g., you cannot know exactly how many people scored 90)

Grade f(x)
0-59 5
60-100 85
Too many classes
X f(x)
Grade Frequency Optimal
(3-7 classes)
58 1

59 1 X f(x)
60 2
Grade Frequency Too few classes
50 - 59 2
61 1 X f(x)
60 – 69 8 Grade Frequency
62 0
70 – 79 14 0 – 59 5
63 1
80 – 89 23 60 - 100 85
64 2
90 - 100 18
65 0

66 0

67 1

68 1

69 0

70 3

71… 1
 Grouped Frequency Table
❖ Grouped frequency – data is presented in classes of values.

❖ Each value will appear in only one class (category) so that they are

– mutually exclusive (the boundaries do not overlap) and

– collectively exhaustive (no values are left unclassified).

Grades f(x) Grades f(x) Grades f(x)

70 – 79 45 70 – 80 45 70 – 79 45

80 - 89 33 80 - 90 33 81 - 90 33

mutually exclusive mutually exclusive mutually exclusive

collectively exhaustive collectively exhaustive collectively exhaustive
 Grouped Frequency table:
Continuous Variable
 The categories of a continuous variable must cover the entire range of
values without leaving empty spaces (unrepresented values).

X f(x)
Age Frequency
21 - 35 95

36 - 50 57 ? 35.8

 Solution: Define the exact limits of each class by calculating half the
distance between the upper score limit of a class and the lower score
limit of the following class.
Exact Limits and Score Limits

Classification Presentation

Exact limits Score limits f(x)

Frequency
20.5 - 35.4999… 21 - 35 95

35.5 – 50.4999… 36 - 50 57
 Example:
Frequency and
Grouped Frequency
tables
Graphical Representation of Data
 Graphical representations allow us to quickly evaluate a large
data collection and notice the distribution's key properties.

 An important factor in determining the form of graphical

representation is the scale by which the variable is measured.

Scales of measurement Graphical representation

Nominal Pie chart / Bar Chart

Ordinal Bar Chart

Quantitative- few values Bar Chart

Quantitative– many values Histogram / Polygon

 Pie Chart - Nominal Variables
• The area of the circle includes 100% of the observations

• Each category has its relative portion out of the circle’s area

(slice of the pie).

Pie Chart: Thoughts

during Zoom meetings
 Bar Chart – Nominal, Ordinal or
Quantitative Discrete with few values
x-axis bar corresponds to values
y-axis bar height corresponds to the frequency or relative frequency
(percentage) of the value

Nominal Ordinal

60%
40%
P(x)
F(x)

20%
0%
BA MA PHD
Histogram – Quantitative Discrete
with many values or Continuous
❖ Each category is presented in its exact limits (x-axis).

❖ When categories are of equal width, the height of the bar

represents the number of cases included in each category (y-axis).
Frequency or percentages

Continuous
Discrete variables
variables
Histogram – Quantitative Discrete
with many values or Continuous
Salaried employee in Israel 2012
30%

25%

20%

15%

10%

5%

0%

0
Polygon – quantitative discrete with
many values or continuous
Smoothing the bars. A polygon is sketched by drawing straight lines
between the midpoints of the classes.
30%

25%

20%

15%

10%

5%

0%

0
Ogive - Cumulative Frequency Curve
 The exact limits of each category will be indicated on the x-axis.
Cumulative frequency/percent will be indicated on the y-axis.

The ogive either increases or remains

constant (parallel to the x-axis) but never
X f(x) CF(x) CP(x)
decreases. Number of Frequency Cumulative Cumulative
children Frequency Percents
100%
0 20 20 10%
80%
1 30 50 25%
60%
2 50 100 50%
40%
3 60 160 80%
20%
4 40 200 100%
0%
Total: N=200 N=200 100%
0 1 2 3 4
Ogive - Cumulative Frequency Curve
Cumulative number of products sold over two weeks.
No. of products

Days
Summary of Tables and Charts
Scale Table Chart

Nominal Frequency Pie chart

Qualitative Bar chart

Ordinal Frequency Bar chart

Ogive (cumulative)

Quantitative Quasi-interval Frequency Bar chart

- discrete /interval / ratio Ogive (cumulative)
- Few values

Quantitative Quasi-interval Grouped Histogram

-discrete / /interval / ratio frequency Polygon
continuous - Many values Ogive (cumulative)
Types of Curves: Shapes of Frequency
Distributions
When we have a frequency distribution of a large number of values of a
continuous variable, we can connect the top of the histogram’s bars with
curves (instead of straight lines), so that the frequency distribution is
illustrated more accurately.

0.125

0.100
Relative Frequency

0.075

0.050

0.025

0.000
53 55 57 59 61 63 65 67 69 71 73 75 77
Height in inches (midpoints)

Many types of continuous distributions are different in uniformity,

the level of symmetry (skewness).
Uniform Distribution (Rectangular)

X f(x)
Age Frequency

20-25 20

25-30 20

30-35 20

35-40 20

40-45 20 20 25 30 35 40 45

A uniform curve is created when the frequency of cases in

each class is equal.
Symmetrical Curves – Bell Shaped
0.125

0.100
Relative Frequency
0.075

0.050

0.025

0.000
53 55 57 59 61 63 65 67 69 71 73 75 77
Height in inches (midpoints)

Symmetrical, bell curve distributions (such as normal distributions) - most

observations are located in the center of the distribution, and as you move
away from the center they gradually become less concentrated.
The distribution on both sides of the bell is symmetrical.

Example: Height distribution: most are around average height (middle of

the distribution), and fewer are very tall or very short.
Symmetrical Curves – Bimodal (U shaped)

Symmetrical distribution - most observations are located in the extremes

of the distribution, and as you move towards the center, they gradually
become less concentrated.

Example: Height distribution of combined 1st grade and 6th grade classes.
Asymmetrical Curves – Positively skewed

Most observations are concentrated around the lower values, and they
gradually become less concentrated as you approach the higher values.

Example: Income distribution: Most households have low income, and

few have high income.
Asymmetrical Curves – Negatively skewed

Most observations are concentrated around the higher values, and they
gradually become less concentrated as you approach the lower values.

Example: Life expectancy distribution: Most people die at an old age

(around 85-90) and few die at a young age.
 Skewness

http://www.youtube.com/watch?v=gzSbAkZE8jw