Relationship Analysis

download Relationship Analysis

of 4

Transcript of Relationship Analysis

  • 8/2/2019 Relationship Analysis

    1/4

    Relationship Analysis

    In business or business related situations multiple variables with respect to customers or

    companies exist it becomes critical to understand, establish the relationship between the

    variables for example a relationship between amounts spent on advertisement and sales could be

    critical.

    The relationship between income, age of the customers and bill amount per visit could be critical

    to a chain of retail.

    Relationship analysis when the data is qualitative:

    When the data is qualitative cross tabs, Chi-Square Tests are used primarily to understand the

    relationships.

    For example to understand the relationship between occupation and defaulting nature of

    customers one can use Cross Tabs and Chi-Square Tests.

    When the data is quantitative or numeric:

    When the data is quantitative relationship analysis is performed by:

    1. Scatter Diagram2. Karl Pearsons Coefficient of Correlation or Product moment correlation.3. Spearmans Rank Correlation or Rank Correlation4. Regression

    Represent the following data as a Scattered diagram and comment on the Correlation.

    X Y

    10 60

    15 6419 71

    21 78

    25 82

    30 85

    32 90

    35 91

  • 8/2/2019 Relationship Analysis

    2/4

    From the scatter Diagram One can observe that there is +ve correlation in the data.

    Which indicates that behavior of X & Y is same i.e. either both increasing or both decreasing.

    Identifying correlation by scatter Diagram:

    Scatter Diagram for a data is obtained by taking values of X on X-Axis and of Y on Y-Axis, the

    correlation is identified as follows:

    1. Perfect positive correlation: Two Variable X & Y said to have perfect +ve correlation ifboth increasing or both decreasing in proportion. The behavior of both the variables is

    same.

    If variables have Perfect positive correlation then their scatter diagram is as follows:

    0

    10

    20

    30

    40

    50

    60

    70

    80

    90

    100

    0 10 20 30 40

    Y

    Y

    Linear (Y)

    0

    1

    2

    3

    4

    5

    6

    7

    0 2 4 6 8 10

    y

    y

  • 8/2/2019 Relationship Analysis

    3/4

    2. Positive Correlation: Two variable X & Y are said to have +ve correlation if oneincrease then the other also increases, if one decreases other also decreases. If two

    variables have Positive correlation then their scatter diagram is as follows:

    3. Perfect Negative Correlation: Two variable X & Y are said to have -ve correlation ifone increase the other decreases if one decreases the other increase in propotion. If two

    variables have Perfect Negative correlation then their scatter diagram is as follows:

    0

    10

    2030

    40

    50

    60

    70

    80

    90

    100

    0 10 20 30 40

    Y

    Y

    Linear (Y)

    0

    1

    2

    3

    4

    5

    6

    7

    0 2 4 6 8 10

    y

    y

  • 8/2/2019 Relationship Analysis

    4/4

    4. Negative Correlation: Two variable X & Y are said to have -ve correlation if oneincrease the other decreases if one decreases the other increase. If two variables have

    Perfect Negative correlation then their scatter diagram is as follows:

    If the scatter diagram is none of the above then data said to have no correlationthe

    scatter diagram mabe as follows:

    0

    1

    2

    3

    4

    5

    6

    7

    0 2 4 6 8 10

    y

    y