If two samples with unknown population distribution test whether the two samples have the same distribution, a nonparametric test of two independent samples is used to test whether there is a significant difference between two independent samples drawn from different populations, null hypothesis: two independent samples or population distribution have no significant difference.
Without further ado, directly on the manipulation.
raw data
Question: A two-sample independent test determines whether there is a significant difference between the number of shots shots from two groups
Action: Analyze → nonparametric test → 2 independent samples → the old dialog box
2 independent sample operations
List of test variables: Number of shot hits
Grouping variables: groups, defining groups (1, 2)
The type of inspection
Mann-Whitney U: Testing whether two samples are equal in position as a whole is equivalent to a sum test (the most extensive) of ranks on two samples
Kolmogorov-Smirnov Z: It is based on the maximum absolute difference in the cumulative distribution of two samples, when the difference is large, the two distributions are treated as different distributions, and test whether the two samples have significant differences in unknown shape
Moses limit reaction: assuming that experimental variables are in one direction, affecting some subjects in the opposite direction affecting other subjects, this method is to reduce the influence of extreme values, control the initial span of the sample, and measure the degree of influence of the extremums in the experimental group on the experimental span, because unexpected outliers may easily deform the span, so after excluding the maximum and minimum values of each 5%, compare whether the difference between the two samples is equal
Wald-Wolfowitz run: A run test that combines or sorts data from two samples, and if the two samples are the same population, then the two groups should be randomly spread across the grades, which is a rank test
Options → descriptive, quartiles
Output the results
<col>
Descriptive statistics
N
mean
standard deviation
Minimum value
Maximum value
Percentile
25th
50th (median)
75th
Number of shots hit
40
5.50
2.918
1
10
3.00
5.00
8.00
Constituencies
1.50
.506
2
The table above shows 40 shooting hits, with an average of 5.5 and a standard deviation of 2.918.
Mann-Whitney U test
order
Rank mean
Rank and order
20
19.90
398.00
21.10
422.00
total
Test statistic a
Mann-Whitney U
188.000
Wilcoxon W
With
-.327
Asymptotic significance (bilateral)
.744
Precise significance [2* (unilateral significance)]
.758b
a. Grouping variables: Groups
b. No amendments were made to the junction.
The above table shows that the number of cases in group 1 and group 2 is 10, the mean value of group 1 is 19.90, the average of group 2 is 21.10, and the asymptotic significance (bilateral) is 0.744>0.05, so the null hypothesis cannot be rejected, indicating that there is no significant difference in the number of shots hits between the two groups.
Moses inspection
frequency
1 (Control)
2 (Test)
Test statistics a,b
Control the group observation span
38
Significance (unilateral)
.500
The trimmed control group span
Outliers trimmed from each end
a. Moses inspection
b. Grouping variables: Groups
As can be seen in the table above, the revised significance (one-sided) is 1.000>0.05, indicating that the null hypothesis that there is no significant difference in the number of shots hits between the two sets cannot be rejected.
Two-sample Kolmogorov-Smirnov Z test
The most extreme difference
absolute value
.100
correct
negative
-.050
Kolmogorov-Smirnov Z
.316
As you can see from the table above, the asymptotic significance (bilateral) is 1.000>0.05, indicating that the null hypothesis is not rejected, that is, there is no significant difference in the number of shots hits between the two sets.
Wald-Wolfowitz test
The number of Runs
Asymptotic significance (unilateral)
Minimal possible
9c
-3.684
.000
Maximum possible
33c
4.005
a. Wald-Wolfowitz test
c. There were 7 intergroup knots involving 33 cases.
As can be seen from the table above, the minimum number of runs (minimum probability) is 9, the maximum number of runs (maximum probability) is 33, and the asymptotic significance (unilateral) of the minimum number of runs is 0.000<0.05, indicating that the acceptance of the null hypothesis, that is, there is a significant difference in the number of shot hits between the two groups, and the asymptotic significance of the maximum number of runs (unilateral) is 1.000 >0.05, indicating that the null hypothesis cannot be rejected, that is, there is no significant difference in the number of shots hits between the two groups.
Taken together, all four tests show that the null hypothesis cannot be rejected, that is, there is no significant difference in the number of shots hits between the two sets.
Today's data analysis is learned here, there are any questions can comment on the message, if you want to see the operation of the explanation, you can private message me. Thank you all for your likes, attention and retweets.