Chapter 6

TEST THE RELATIONSHIP
One of the techniques of inferential statistics that is widely used in field is meant to test
for the relationship between some symptoms. It is because people see the emergence of the
symptoms the symptoms of a certain disruptions with symptoms other symptoms. The
symptoms the symptoms visible related can be observed, tested, measured, and
dikuantifikasikan so that the existence of the relationship can be tested the truth through
statistics test. The concept of the premise in this issue is whether the high score low or the
occurrence of a variable will be followed by high score low or the occurrence of the other
variables. The way that should be taken to perform a test of both the ability to a number of
samples taken random. The results of field testing was then tested with the technique of the
statistics in the form of the relationship test technique, or known with the title correlation
technique. The statistics are intended to test is whether or not the relationship between the
symptoms including inferential statistics. If that been correlated is between the data that the
scale ordinal scale, correlation technique that used is the correlation between good governance
levels. If that dikorelasikan is between the data that the scale interval scale with nominal scale,
correlation technique that is used is the point-biserial correlation. Correlation tests involving
two variables is often referred to as a bivariate correlation. The correlation is done for many
groups called as a correlation between the variables. If the correlation between the variables
controlled this will be a partial correlation. Conversely if one dikorelasikan variable with some
variables, correlation is called as a double correlation.
A. PRODUCT-MOMENT CORRELATION
The concept of test thought the relationship is high low a score variables will be
followed systematically by high score low the other variables theoretically have
relationship characteristic. Statistical tests can actually includes a review of the
relationship and the predictive test. The data can be described through the scatter diagram
that shows the relationship between both variables score. Correlation coefficient (r) ranged
between -1,00 until hiehest both positive or negative shows that there is a perfect
relationship. R coefficient perfect positive occurs if the level of each leaderboard on a
variables will be followed consistently and systematically by the level of the other
variables. R coefficient perfect negative will happen if there is a reversal of the high score
low in a variable on other variables. If there is no completely kesistematisan between both
variables which dikorlasikan, it shows that there is no correlation between both of them.
The formula to calculate the correlation coefficient (R) is:
Or can use the formula of the deviation and kovarian:
The use of the two equations above more quickly and compact when compared with
the other formula.
To view the value of the critical value r Product-Moment, first must be determined great
degree of freedom (db), which is with the formula: db = N-1.
B. GOVERNANCE CORRELATION LEVELS
Governance correlation levels used to mengorelasikan between the two groups of

data that shows the sequence of the elementary, or is the data periodically ordinal. There
are two kinds of the formula in governance levels, namely:
1. Governance Levels Spearman correlation (rho)

In the equations above numbers 1 and 6 is a constant numbers
To test the rank of the significance of governance levels Spearman correlation (rho)
done Table consultation Rho values that are on the attachment.
To work out the calculations intended in the above formula, it must first be:
1. Ranking for achievement indexes to rank 1,2,3, and so on.
2. Calculates the difference between the two ranks that yields D (difference), and
then squared that produces D2.
2. Kendall Correlation Levels (tau)

The correlation of the Kendall (tau) level structure is often used interchangeably
with Spearman's (rho) level correlation as found previously. The formula used to
calculate Kendall's correlation coefficient is:
𝐏−𝐐
𝐭𝐚𝐮 =
𝐍(𝐍 − 𝟏)
𝟐
P : Higher number of rank numbers
Q : Lower number of rank numbers
N : Total of the members
(Note: Notice how to calculate the number of higher and lower ranking figures)
To rank and calculate higher and lower number of numbers, please follow the steps:
1. Ranking for the first assessment group.
2. Ranking for the second assessment group.
3. Counting of higher ranked and lower ranked contestants is done one by one per
club for second ranking.
4. The next count for the top ranked club.
If in a ranking example, if in the rankings there are twin ranks or more than one,
there is a correction factor formula. In other words the previous formula only applies
fully if there are no more than one rank, and no more than one rank, the formula to be
corrected. Corrected formulas:
𝑷−𝑸
𝒕𝒂𝒖 =
𝑵(𝑵 − 𝟏) 𝒁𝟏 (𝒁𝟏 − 𝟏) + ⋯ + 𝒁𝟏𝒏 (𝒁𝟏𝒏 − 𝟏) 𝑵(𝑵 − 𝟏) 𝒁𝟐 (𝒁𝟐 − 𝟏) + ⋯ + 𝒁𝟐𝒏 (𝒁𝟐𝒏 − 𝟏)
√[ − ][ − ]
𝟐 𝟐 𝟐 𝟐
Z1 : Rankings of the First Twin

Z2 : Ranking of the Second Twin
C. CORRELATION BUSINESS POINTS

Used if the competition data to be sought in the form of interval scale data and
nominal scale dichotomous. Nominal scale data does not have the characteristic of normal
distribution as well as interval-scale data. As for the biserial point correlation formula:
̅𝐩 − 𝐗
𝐗 ̅𝐪
𝐫𝐩𝐛𝐢 = √𝐩𝐪
𝐬
rpbi : Coefficient of correlation of biserial points sought

̅p
X : The average arithmetic interval data is categorized dichotomy 1
̅
Xq : The average arithmetic interval data is categorized dichotomy 0
s : Standard deviation from the entire interval data

p : Proportion of cases dichotomized 1
q : Proportion of cases dichotomized 0

Chapter 6

Uploaded by

Copyright:

Available Formats

Chapter 6

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter 6

Uploaded by

Copyright:

Available Formats

TEST THE RELATIONSHIP

The formula to calculate the correlation coefficient (R) is:

Or can use the formula of the deviation and kovarian:

B. GOVERNANCE CORRELATION LEVELS

Governance correlation levels used to mengorelasikan between the two groups of

1. Governance Levels Spearman correlation (rho)

2. Kendall Correlation Levels (tau)

Z1 : Rankings of the First Twin

C. CORRELATION BUSINESS POINTS

rpbi : Coefficient of correlation of biserial points sought

s : Standard deviation from the entire interval data

You might also like