Identify Rules that Predict Patient’s Heart Disease - An Application of Decis...
Correlation
1. Correlation
The concept of correlation defines as the degree to which two or more variable are
interrelated. The durability of positive correlation differs from -1.00 to +1.00. The correlation of
-1.00 defines “Perfect Negative correlation” as one unit increases in one variable other variable
tends to be decrease. Whereas, the correlation of +1.00 define as “Perfect Positive correlation” in
which one unit increase in one variable and other variable also increases. There are three types of
correlation
1. Positive/ Direct Relationship
2. Negative/ Indirect/Inverse Relationship
3. No Relationship
TWO HYPOTHESES:
H0: There is no relationship between two variables
HA: There is the relationship between two variables
PURPOSE OF CORRELATION ANALYSIS
To form strength and order of relationship between variables
To form a degree in which two variable interchange with each other
Correlation apply in following fields:
i. Scientific Data Analytic
ii. Marketing research
iii. Business Analysis
2. Test Statistic: Correlation
To calculate Average of each variable in SPSS we have to open the dialog box in
Transform Select Compute Variable
Fig: 1.1
3. In the Compute Variable window, Calculate Average of each variable BE, PQ, BL, BA
and Bas as shown in fig: 1.2
Fig: 1.2
After calculating Average of each variable, the SPSS output generate all variable been
executed as you can see in fig 1.3
Fig: 1.3
4. To calculate Bivariate we have to open dialog box
Analyze Correlate Bivariate
5. Bivariate Correlations box will open
Select Average variable and Shift two variable BE and PQ in to Variable Box
In Test Significance Box Select Two- tailed because we do not have any presumption
whether it is positive or negative among two variable BE and PQ
Under Correlation Coefficients Box “Pearson” is mark off by default
We also leave default check on “Flag significant correlation”
6. INTERPRETATION:
The above table shows the result of “Correlation Analysis”. In this table BE and PQ have
positive relationship with the value of .410% which mean that there is strong relationship among
BE and PQ. The total numbers of respondent acknowledge both items are 200. Based on
significance we reject null hypothesis and accept alternative hypothesis because our P-value is
lesser than
“N” represents number of cases
which is used in Correlation Test as;
we do not have any missing data in
our data. Our Correlation analysis is
based on 200 numbers of cases.
Based on Significance Value, Our
hypothesis is
H0: BE and PQ has no relationship
HA: BE and PQ has the relationship
Test Statistic:
Correlation Bivariate/
Pearson
Pre-determined Value
5%