Scribd
Upload
1K views

Sign Test

The sign test is a nonparametric test that can be used to determine if two groups are equally sized when the data is not normally distributed or variances are not equal. There are two types of sign tests: 1) a one sample sign test that tests if positive and negative signs are equally likely and 2) a paired sample sign test, also known as an alternative to the paired t-test, that tests if the signs of differences between pairs are equally likely. The sign test for paired data follows similar logic as the paired t-test. To perform a sign test, values are assigned a positive or negative sign depending on if they are above or below the mean. The number of positive and negative signs are counted and compared

Uploaded by

prasadkulkarnigit
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
1K views

Sign Test

The sign test is a nonparametric test that can be used to determine if two groups are equally sized when the data is not normally distributed or variances are not equal. There are two types of sign tests: 1) a one sample sign test that tests if positive and negative signs are equally likely and 2) a paired sample sign test, also known as an alternative to the paired t-test, that tests if the signs of differences between pairs are equally likely. The sign test for paired data follows similar logic as the paired t-test. To perform a sign test, values are assigned a positive or negative sign depending on if they are above or below the mean. The number of positive and negative signs are counted and compared

Uploaded by

prasadkulkarnigit
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 5

Sign Test

While using statistical methods, the data used must be formatted and hence different
assumptions are made on them. This assumption is only good when we are doing a clinical
test. While dealing with raw data, it will be important to apply a suitable transformation. Here
comes the problem that every transformed data may not satisfy the assumption made and
hence using the traditional parametric method of analysis becomes inappropriate. For this
reason. The sign test is defined under the heading of non parametric test, which can only be
used if the observed differences between any two conditions are significant. It tests the null
hypothesis and checks if two groups are equally sized. It means by the use of the sign test, it
will be possible to determine when variable is greater than the other.

Types of Sign Test


Back to Top

The sign test is used to test the null hypothesis that two groups are of equal size. Many of the
hypothesis tests require normal distributed populations or some tests require that population
variances be equal.statisticians have developed hypothesis tests that are 'distribution free'.
Such tests are called nonparametric tests. One of the easiest nonparametric tests to perform is
the sign test.

Two types of sign test can be explained as follows:

1. One Sample: Here, the hypothesis is formed so that positive(+) and negative(-) signs
are the values of random variables which will have equal size.

2. Paired Sample: This test also known as alternative to the paired sample t-test. Here,
the signs are in paired sample tests or in the form of before-after study. Here, null
hypothesis will be set up so that the sign are of equal size or their population means
are equal to the sample mean.

Sign Test for Paired Data


Back to Top

A nonparametric paired data are taken in the same logic as the parametric paired t-test. If the
distribution of the difference are non normal and cant be normalized by any transformation,
then a 1 sample nonparametric test will be appropriate. Here, there are mainly two choices:

1 sample sign test

Wilcoxon sign rank test, each with their own assumption.

These test makes it possible to compare observed and hypothesized medians.


Process of Sign Test

Step 1:

Consider a distribution. Then, calculate its + as well as sign. Plug in a positive (+) sign for
those values which has greater than the mean value and put a negative () sign for those value
which has less than the mean value. Plug in zero when the value becomes equal to the mean
value.

Step 2:

Denote n as the total number of signs excluding zero sign and the s by the number of less
frequency signs. If the sample is very small, use the formula to get the critical value at 5% of
the significance level.

Step 3:

K = n12
- 0.98n

When the sample size is large, we use another formula, as follows.

Z = Snpnp(1p)

Under the condition when p = 12

We use the distribution formula for binomial = nCx qn - x px

Step 4:

Next, the values of S and the critical value K are compared.

If s > K, the null hypothesis is accepted.


If s < K, the null hypothesis is accepted.

If the sample size is very large, S is compared with the Z value. For the rest of the calculation
it will be very helpful to use the SPSS software.

Sign Test Table


Back to Top

Critical values for Sign Test are given in the table shown below. Left values of s give P(S < s)
and right values of s give P(S > s).
Sign Test Example
Back to Top

Given below are some of the examples on Sign Test.

Solved Example

Question: A test is done before and after training. About 15 people were tested and it has
been seen that the datas are not normally distributed. We need to find whether the training has
improved the test scores.

The following are the raw data:


Person Before score After score After before Sign change
1 5 6 1 +
2 3 2 -1 -
3 4 4 0
4 2 4 2 +
5 1 3 2 +
6 6 6 0
7 7 7 0
8 3 5 2 +
9 2 3 1 +
10 3 5 2 +
11 5 5 0
12 1 3 2 +
13 4 4 0
14 4 5 1 +
15 3 2 -1 -

Solution:

Step 1:
Using above table
Total number of people = 15
Number of positive sign = 8
Number of negative sign = 2

The binomial test tests the probability of k when its greater than or equal to 8 has positive
changes,
Here the probability p = 12
for positive and

the probability q = 12 for negative.

Step 2:
Now lets take the Hypothesis:

H0: There are less than or equal positive as negative changes (p 0.5)

H1: There are more than positive as negative changes (p > 0.5)

Now lets take a small sample of 10 members, so n =10


p = 0.5 and q = 1 - p = 0.5

Step 3:
The probability for the events k equal or greater 8 is given by using the binomial formula

ni=k(ni) pi qn - 1

= (108)(12)8(12)2 + (109)(12)9(12)1 + (1010)(12)10(12)0

= 0.044 + 0.0097 + 0.00097

= 0.05467

Step 4:
Lets check the one sided distribution

The probability (x 8) = 0.0547.

The significance level taken is 5 percent. So the probability is greater than the 5 percent
significance level. Hence H1 is discarded. The training doesnt show any improvement in test.

You might also like