Chi Square Test Excel Template

Download as xlsx, pdf, or txt
Download as xlsx, pdf, or txt
You are on page 1of 18

ChiSquare Test

Chi Square Test is a test of the validity of a hypothesis.


The Chi Square P Value tells us if our observed results are statistically significant or not.
A statistically significant result means that we reject the null hypothesis (the null hypothesis in statistics is a statement or hypo
A Chi-Square P-Value is a number between 0 and 1. A Chi-Square P-Value less than 0.05 usually lead to rejection of the null hy

The CHISQ.TEST function in Excel returns the chi-squared probability of something across datasets. The
CHISQ.TEST(actual_range, expected_range)

sample dataset of dress styles worn by men. So the manufacturing company decided to measure the p val
Here sales are in thousands.

Apparel Trouser Shorts Shirts T-Shirts Socks Shoes Belts Caps


Observed Sale 24 22 19 21 18 19 20 23
Actual Sales 20 20 20 20 20 20 20 20

P Value Calcuation 0.970076

the test returned around 0.97 or 97% relation between the two, it means the observed sales is close to the expected sales res
The test returning value above 0.5 shows the good relation and indicates the level of dependence between the two whereas t

1. The function only works with numbers. If any argument other than cumulative is non numeric, the function considers the no
2.Value in decimal or value in percentage is the same value in Excel. Convert the value to percentage, if required.
3.The function returns #N/A error if the length of the two range arguments is not equal or the dataset contains only one value
4.The function returns #DIV/0 error if any of the values provided in expected_range is zero
5.The function returns #NUM! Error, if any number in range is negative.
tatistics is a statement or hypothesis which is likely to be incorrect).
ead to rejection of the null hypothesis.

ng across datasets. The function takes the actual dataset and expected dataset as argument and returns the probab

ided to measure the p value relation between the observed sales and actual sales.

lose to the expected sales result.


e between the two whereas the test returning value less than 0.5 is an indicator of independence.

, the function considers the non numeric value as 0 and returns the result.
tage, if required.
ataset contains only one value i.e the length and width equal to 1.
nt and returns the probability relation between the two.
Observed Hall A Hall B Hall C Hall D Total
Chairs 100 92 89 81 362
Tables 100 77 78 85 340
Stools 50 81 75 76 282
Total 250 250 242 242 984

Expected Hall A Hall B Hall C Hall D Total


Chairs 91.97154 91.97154 89.02846 89.02846 362
Tables 86.38211 86.38211 83.61789 83.61789 340
Stools 71.64634 71.64634 69.35366 69.35366 282
Total 250 250 242 242 984

Chi Square Value 0.031373


Actual Hours Expected
Analyst Completion % Chi Square Value
Datasets Worked Datasets
Rohit 5 2 100 9.04 0.0204648696745034
Susan 14 7 100 9.04
Sharon 12 5 90 9.04
Monica 10 10 100 9.04
Ralph 6 5 90 9.04
Puneet 4 4 84 9.04
Jacques 14 12 87 9.04
Supriya 15 12 92 9.04
Archana 8 6 78 9.04
Chen 6 5 89 9.04
Ian 5 4 100 9.04
Tsetou 7 5 79 9.04
Mark 2 1 50 9.04
Richa 11 10 100 9.04

10 10
9 11
12 12
9 6
11 7
15 14
12 11
11 11
4 3
5 3
99 9.04
98 9.04
100 9.04
100 9.04
93 9.04
100 9.04
99 9.04
100 9.04
100 9.04
97 9.04
T.TEST is used to determine the confidence of an analysis.
Mathematically, it is used to know if the mean of the two samples are equal or not. T.TEST is used to accept or reject the null h

he T TEST or Student’s test is used to determine the confidence of an analysis. In mathematical term, it is used to know if the m
(Null hypothesis states that there is no significant difference in mean due to a given reason between two gr
Let’s say, we do a survey on two groups. In the end, we get the mean of these surveys and conclude that group 1 has done be

T.TEST(group1, group2, tails, type)

Group1: the first group of experiments.


Group2: the second group of experiments. It can have the same subject as of group1, different subjects wit
Tails: it has two options, 1 tailed, and 2 tailed. 1 tailed T TEST is used when we know that the effect of the

Types: 
There are three options available (1,2,3). 
1. 1 (Paired): we use it when group1 and group2 have same subjects. 
2. 2 (Two sample equal variance): When the subjects of these groups are different but variance is same.
3. 3 (two sample unequal variances): When subjects of two groups are different and variance is also different.
ed to accept or reject the null hypothesis.

erm, it is used to know if the mean of the two samples are equal. T.TEST is used to accept or reject the null hypothesis.
en reason between two groups. )
ude that group 1 has done better from group 1  due to a reason. But how confident we are, that it is not just a fluke but ther

up1, different subjects with the same variance and different subjects with different variance.
now that the effect of the experiment is one-directional. 2 tailed T TEST it is used when we don't know the direction

nce is same.
e is also different.
hypothesis.

is not just a fluke but there’s a large possibility of it. This is determined by T.TEST.  

don't know the direction of the effect of the experiment. 


Study with Break Study Without Break
Group A Group B
1 82 80
2 85 82
3 82 83
4 87 81
5 79 79
6 82 85
7 92 90
8 87 88
Meanq ??? ???
 To test this I conducted an experiment between two groups. Group 1 studied with break and group 2 stu

Null Hypothesis (H0): There is no difference in score due to studying with Break.
Assume that the subjects of these two groups are the same. We also assume that the effect of this experiment will be one-dire
T.TEST confidence level is between 0% to 50%. This is too low. Hence we accept the null hypothesis.
0.175308

Two Tailed T Test 0.350616662820207


Similarly for 2 tailed T TEST we will write this formula
T.TEST(B3:B10,C3:C10,2,1)

This will result 0.350. Which double of 1 tailed T.TEST


h break and group 2 studied without break. We collected the data and calculated the mean in Excel. The average s

is experiment will be one-directional. We will write this formula to get 1 tailed T-TEST with the paired group.
t the null hypothesis.
n in Excel. The average score of  group 1 is slightly greater than group 2. 
The F.TEST Function is used to calculate F statistic of two samples in excel internally and returns the two
nally and returns the two tailed probability  of the F statistic under Null Hypothesis.
DEVSQ function is a built-in statistical function to calculate the sum of squared deviations from the mean o
eviations from the mean or average of the range of data values provided.
Calculate the Z score for the normal cumulative distribution for the pre specified values using the NORMDI
alues using the NORMDIST function in Excel.

You might also like