Spearman's Rho

Kendall's Tau
Phi Coefficient
Cramer's V

BSN IV – 2 GROUP 4
SPEARMAN'S RHO
• Spearman's rank correlation coefficient or Spearman's rho,
named after Charles Spearman and often denoted by the Greek
letter ρ (rho) or as rs, is a non-parametric measure of statistical
dependence between two variables.

• It assesses how well the relationship between two variables can

be described using a monotonic function.

• If there are no repeated data values, a perfect Spearman

correlation of +1 or −1 occurs when each of the variables is a
perfect monotone function of the other.

• Measure the linear relationship between two variables. It differs

from Pearson's correlation only in that the computations are done
after the numbers are converted to ranks.
 
• The sign of the Spearman correlation indicates the direction
of association between X (the independent variable)
and Y (the dependent variable).

• If Y tends to increase when X increases, the Spearman

correlation coefficient is positive.

• If Y tends to decrease when X increases, the Spearman

correlation coefficient is negative.

• Oftenly described as being "nonparametric."

PROCEDURE FOR USING SPEARMAN'S RANK CORRELATION

1. State the null hypothesis i.e. "There is no relationship between the

two sets of data.“

2. Rank both sets of data from the highest to the lowest. Make sure to
check for tied ranks.

3. Subtract the two sets of ranks to get the difference d.

4. Square the values of d.

5. Add the squared values of d to get Sigma d 2.

6. Use the formula Rs = 1-(6Sigma d2/n3-n) where n is the number of

ranks you have.
 EXAMPLE

Problem:  

Six nursing students' have the following rankings

in Clinical performance and patient satisfaction
on the care. Is there any association between
their clinical performance and patient’s
satisfaction on the care?
Respondent Clinical Patient
performance satisfaction
on the care

1 85.36 89.67

2 82.98 86.71

3 87.76 88.18

4 84.50 85.63

5 84.03 84.56

6 86.33 87.40
Respondent Clinical Patient satisfaction on
performance the care

1 3 1

2 6 4

3 1 2

4 4 5

5 5 6

6 2 3
Respondent Clinical Patient D1
performance satisfaction on
the care
1 3 1 2
2 6 4 2
3 1 2 -1
4 4 5 -1
5 5 6 -1
6 2 3 -1
N= 6
Respondent Clinical Patient D1 D2
performance satisfaction
on the care
1 3 1 2 4
2 6 4 2 4
3 1 2 -1 1
4 4 5 -1 1
5 5 6 -1 1
6 2 3 -1 1
N= 6 12
 Solution:

= 1- 6(12)/ 6(36-1) = 1-72/210 = 1 - 0.34 = 0.66

1. If the Rs value...
... is -1, there is a perfect negative correlation.
...falls between -1 and -0.5, there is a strong negative correlation.
...falls between -0.5 and 0, there is a weak negative correlation. 
... is 0, there is no correlation
...falls between 0 and 0.5, there is a weak positive correlation.
...falls between 0.5 and 1, there is a strong positive correlation
...is 1, there is a perfect positive correlation
between the 2 sets of data.

2. If the Rs value is 0, state that null hypothesis is accepted. Otherwise,

say it is rejected.
KENDALL'S TAU
• Kendall rank correlation coefficient, commonly referred to
as Kendall's tau (τ) coefficient, is a statistic used to measure
the association between two measured quantities. It was
developed by Maurice Kendall in 1938.

• A tau test is a non-parametric hypothesis test which uses the

coefficient to test for statistical dependence. Specifically, it is a
measure of rank correlation: that is, the similarity of the orderings
of the data when ranked by each of the quantities.
ADVANTAGES OF KENDALL’S TAU :
• The distribution of Kendall’s tau has better statistical
properties.

• The interpretation of Kendall’s tau in terms of the probabilities

of observing the agreeable (concordant) and non agreeable
(discordant) pairs is very direct.

• In most of the situations, the interpretations of Kendall’s tau

and Spearman’s rank correlation coefficient are very similar
and thus invariably lead to the same inferences.

• In Spearman’s rank correlation coefficient, the measure of rank

correlation is the more widely used rank correlation coefficient.
• Symbolically, Spearman’s rank correlation coefficient is denoted
by rs . Spearman’s rank correlation coefficient is given by the
following formula:

• rs = 1- (6∑di2 )/ (n (n2-1)), here di in Spearman’s rank correlation

coefficient represents the difference in the ranks given to the values
of the variable for each item of the particular data. This formula of
Spearman’s rank correlation coefficient is applied in cases when
there are no tied ranks. However, in the case of fewer numbers of
tied ranks, this approximation of Spearman’s rank correlation
coefficient provides sufficiently good approximations.
Tau-a

Tau-a statistic tests the strength of association of the cross tabulations.

Both variables have to be ordinal. Tau-a will not make any adjustment
for ties.

Tau-b

Tau-b statistic, unlike tau-a, makes adjustments for ties and is suitable
for square tables. Values of tau-b range from −1 100% negative
association, or perfect inversion to +1 100% positive association, or
perfect agreement. A value of zero indicates the absence of
association.

Tau-c

Tau-c differs from tau-b as in being more suitable

for rectangular tables than for square tables.
 Respondent Home Self-care ability
hemodialysis
perception

Kendall's Tau-C was used to examine the strength of association

between the home hemodialysis patients' perceptions and self-care
ability.
 • A hemodialysis patient's perception regarding HHD was significantly
and negatively associated with both social support, Kendall's Tau-C =
-0.288, p = 0.038 and communication, Kendall's Tau-C = -0.353, p =
0.001. This indicates that when a patient scores high on the Patient
Perception Survey, meaning he or she has negative perceptions
regarding HHD, the patient as well has low levels of social support and
communication. The associations between a hemodialysis patient's
perception regarding HHD and all other subscales were not significant.
Table 4 shows the strength of the association between patient
perceptions and each of the six subscales, such as physical,
communication, ability to maintain self-care, social support,
psychological status, and nutritional status.
• This study utilized the Patient Perception Survey to obtain measures
of hemodialysis patient perceptions regarding HHD. Results from
this sample of 49 patients found that 46.9% of the patients identified
negative perceptions of HHD and 53.1% of the 49 patients identified
positive perceptions of HHD.
Kendall's Tau-b

Kendall's Tau-b is a measure of association often used with but not limited to
2-by-2 tables. It is computed as the excess of concordant over discordant pairs (C -
D), divided by a term representing the geometric mean between the number of
pairs not tied on X (X0) and the number not tied on Y (Y0):

Tau-b = (C - D)/ SQRT [(C + D + Y0)(C + D + Y0)]

There is no well-defined intuitive meaning for Tau -b, which is the surplus of
concordant over discordant pairs as a percentage of concordant, discordant, and
approximately one-half of tied pairs. The rationale for this is that if the direction of
causation is unknown, then the surplus of concordant over discordant pairs should
be compared with the total of all relevant pairs, where those relevant are the
concordant pairs, the discordant pairs, plus either the X-ties or Y-ties but not both,
and since direction is not known, the geometric mean is used as an estimate of
relevant tied pairs.
Tau-b requires binary or ordinal data. It reaches 1.0 (or -1.0 for negative
relationships) only for square tables when all entries are on one diagonal. Tau-
b equals 0 under statistical independence for both square and non-square
tables. Tau-c is used for non-square tables.
Tau-c

Kendall's Tau-c, also called Kendall-Stuart Tau-c, is a variant of Tau-

b for larger tables. It equals the excess of concordant over discordant
pairs, multiplied by a term representing an adjustment for the size of the
table.

Tau-c = (C - D)*[2m/(n2(m-1))]

Where:

m = the number of rows or columns, whichever is smaller

n = the sample size.

Hemodialysis patients' perceptions of home hemodialysis and self-care

Author: Visaya, Marie Angela Date published: April 1, 2010

Phi Coefficient

• The phi coefficient is a measure of the degree of association

between two binary variables. It is a nominal association that is
based on the adjusted chi-square.

• In calculating the phi, we divide chi-square by the sample size. The

phi-coefficient was designed for the comparison of truly
dichotomous distributions, i.e., distributions that have only two
points on their scale which indicate some unmeasurable attribute.

• Attributes such as living or dead, black or white, accept or reject,

and success or failure are examples. It is also sometimes known as
the Yule φ Phi coefficient is used for a 2×2 table and a nominal
variable.

• Testing of significance of Phi is the same as chi-square.

FORMULA FOR THE PHI COEFFICIENT
 EXAMPLE
43 persons were asked if they believed that there was any truth in
horoscopes or in the existence of UFOs. The results gave:
Horoscopes

UFO’s Some truth No truth TOTAL

Might exist 14 10 24
Don’t exist 6 13 19
TOTAL 20 23 N=43
• Applying the above formula, φ = 0.266

• This value of φ corresponds to a value of χ2 of 43 × (0.266) 2 = 3.04.

This may then be tested against the relevant value of χ2 for 1 degree
of freedom. An alternative significance test (rarely used) may be
performed by considering the standard error of φ. Calculation of this
is laborious but if N is not too small, then 1/ √N approximates to it.
Interpretation of the Phi coefficient.

I have general rule of thumb for correlation coefficients and you can
use the same rule for the Phi coefficient.

-1.0 to -0.7 strong negative association.

-0.7 to -0.3 weak negative association.
-0.3 to +0.3 little or no association.
+0.3 to +0.7 weak positive association.
+0.7 to +1.0 strong positive association.
ANOTHER EXAMPLE TRY TO COMPUTE:
CRAMER'S V

• Cramer's V is the most popular of the chi-square-based

measures of nominal association because it gives
good morning from 0 to 1 regardless of table size, when
row marginals equal column marginals. It was named after the
Swedish mathematician and statistician Harald Cramér, who
sought to make statistics mathematically rigorous, much like
Kolmogorov's axiomatization of probability theory.

• Significance testing of the Cramer’s V is the same as the chi-

square test where it says that there is a significant relationship
between variables, but it does not say just how significant and
important this is. Cramer's V is a post-test to give this additional
information.
• The value of the Cramer’s V can be reached at 1 if two variables
have an equal marginal. The value of Cramer’s V is always less
than the phi coefficient. The following formula is used to calculate
the value of Cramer’s V:

• Where v is Cramer’s V and n and m are the sample size and time.

• where “min” is a single value, the smaller of the two quantities (r-1)
or (c-1).
• Cramer’s V was calculated as a measure of
the association between IT sophistication for resident care
management and the characteristics of ownership, bedsize, and
location.
 Interpretation: V may be viewed as the association between
two variables as a percentage of their maximum possible
variation.V2 is the mean square canonical correlation between
the variables. For 2-by-2 tables, V = phi (hence some packages
like Systat print V only for larger tables).

 Meaning of association: V defines a perfect relationship as one

which is predictive or ordered monotonic, and defines a null
relationship as statistical independence, as discussed in the
section on association. However, the more unequal
the marginals, the more V will be less than 1.0.

 Symmetricalness: V is a symmetrical measure. It does not matter

which is the independent (column) variable.
  Data level: V may be used with nominal data or higher.

 Other features: V can reach 1.0 only when the two variables
have equal marginals.

 Equal marginals. Cramer's V and all measures which define a

perfect relationship in terms of strict monotonicity require
that the marginal distribution of the two variables be equal
for the coefficient to reach 1.0.

 Following convention, the strength of the relationship will be

interpreted as follows:

 values between 0.0-.30 indicate a weak association.

 values between .31-.60 indicate a moderate association.
 values >.60 indicate a strong association.
THANK YOU FOR LISTENING! ☺☻