The document discusses different types of correlation including positive, negative, simple, partial and multiple correlation. It also defines correlation coefficient and describes methods for studying correlation including scatter diagrams, Karl Pearson's correlation coefficient, rank correlation and concurrent deviation method.
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0 ratings0% found this document useful (0 votes)
28 views2 pages
CORRELATION
The document discusses different types of correlation including positive, negative, simple, partial and multiple correlation. It also defines correlation coefficient and describes methods for studying correlation including scatter diagrams, Karl Pearson's correlation coefficient, rank correlation and concurrent deviation method.
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2
39/2020 STATISTICAL ASSISTANT GRADE II/
STATISTICAL INVESTIGATOR GRADE II TOPIC: CORRELATION
Definition IV. Partial correlation: - Two or more
“Correlation analysis as an attempt to determine variables are recognized and two the degree of relationship between variables”- variables are considered for study by Ya-Lun-Chou keeping other variables as constant. V. Multiple correlation: - Three or more “Correlation analysis deals with the association variables are considered for study. between two or more variables” VI. Linear correlation: - If the amount of -Simpson&Kafka change in one variable follows a constant “Correlation means that between two series or ratio change of other variable. groups of data there exists some casual VII. Nonlinear correlation:- If the amount of relation”.-W.L.King change in one variable does not bear a constant ratio to the amount of change in “When the relationship is of a quantitative other variable. nature, the approximate statistical tool for Degree of correlation discovering and measuring the relationship and I. Perfect positive correlation:- If an expressing it in a brief formula is known as increase(decrease) in the value of one correlation “-Croxton&Cowden variable is followed by the same proportion of increase(decrease) in other Correlation coefficient related variables. Coefficient of correlation is a relative measure II. Perfect negative correlation: - If an showing the degree of relationship between two increase (decrease) in value of one variable. variable is followed by the same proportion of decrease (increase) in other Types of Correlation variable. I. Positive: - If the two variables are moving III. Zero correlation: - If there is no correlation together in same direction then the between variables. correlation is positive Methods for studying correlation II. Negative :- If two variables are moving in I. Scatter diagram opposite direction then the correlation is Scatter diagram is a graphical method of negative studying correlation. it also known as III. Simple correlation: - When only two scatter plot variables are studied it is called simple. Eg:- relationship between Price and Supply of a product Probable error Difference occurring due to taking samples from the means or population II. Karl person’s correlation coefficient(r):- 1−𝑟 2 Pr =0.6745 r( x, y)= 𝑐𝑜𝑣(𝑥,𝑦) 𝜎𝑥 𝜎𝑦 √𝑛
III. Rank correlation
1 To study the correlation between attributes, 𝑐𝑜𝑣 (𝑥, 𝑦) = ∑(𝑥 − 𝑥̅ )(𝑦 − 𝑦̅) Developed by Charls Edward Spearman 𝑛 1 6 ∑ 𝐷2 𝜎𝑥 =√ (𝑥 − 𝑥̅ )2 𝑅 =1− 3 𝑛 𝑁 −𝑁 1 𝜎𝑦 =√ (𝑦 − 𝑦̅)2 IV. Concurrent deviation Method 𝑛 If we want to study the correlation between two series is a The value of coefficient of correlation (r) casual manner and no precision is required. always takes values from -1 to +1 ie; -1≤r≤+1 2𝐶−𝑛 Properties 𝑟𝑐 = ±√± { } 𝑛 1. Correlation coefficient takes any values from -1 to +1 C=number of concurrent deviations 2. i. Magnitude of coefficient of n= number of pairs of observations compared correlation is unaltered if a Note:- constant added (subtracted) 2C>N; r is positive from the value of one variable 2C<N ; r is negative (both). Coefficient of determination ii. Magnitude of coefficient of Coefficient of determination is the square of coefficient of correlation is unaltered if the correlation. Thus coefficient of correlation is r2. It is more value of one variable (both) useful to measure the percentage variation in the dependent are multiplied (divided) by a variable which is accounted for by the independent variable. constant. Eg;- if r=0.7 the coefficient of determination will be 0.49; this 3. Correlation coefficient is symmetric means 49% of the variation in the dependent variable has with respect to variable been explained by the independent variable. ie, r(x, y)=r(y, x) Coefficient of non-determination The ratio of unexplained variance to total variance is called 4. Correlation coefficient between two coefficient of determination. Denoted by k2 ; k2 =1-r2 independent variables is zero. But the converse is not true.