20.10.2011 Statistics

If p < 0,05 then we reject null hypothesis

If p > 0,05 then we not reject null hypothesis

If p = 0,05 then you can choose

t-test

Null hypothesis: samples are equal (are the same)

T-test for Dependent Samples (Desetiboj) Marked differences are significant at p < ,05000

                                               Mean

                                              Std.Dv.

                                                 N

                                                 t

                                                df

                                                 p

                                           points 100 m

871,7179

53,0997

                                         points long jump

803,3333

246,5539

39

1,880152

38

0,067769


Because p (0,06)>0,05, we not reject null hypothesis, which is telling that samples are equal. It
means that from “100m” has obtained same points. Difference is not statistically significant.

From statistical point of view, 871 points from “100 m” is equal 803 point from “long jump”. This
difference is not statistically significant. At the 5% level of statistical significance







                                               Mean

                                              Std.Dv.

                                                 N

                                                 t

                                                df

                                                 p

                                           points 100 m

871,7179

53,0997

                                           points 400 m

725,7949

287,4990

39

3,323692

38

0,001974








Because p (0,00)<0,05, we reject null hypothesis, which is telling that samples are equal. It means
that from “100m” athletes have obtained different  (more) points than from “400 m”. Difference is
statistically significant.


Graph is used to confuse the reader.

There are three lies. Terrible, ugly and statistics.

Do not believe the charts, believe numbers.


Correlation

                                           points 100 m

                                         points long jump

                                           points 100 m

1,0000

,4589

p= ---

p=,003

                                         points long jump

,4589

1,0000

p=,003

p= ---

Because p (0,003)<0,05, we reject null hypothesis, which is telling that correlation coefficient is
ZERO. It means that correlation is not ZERO, and is 0,4589. Correlation is statistically
significant. This relationship is direct.

                                           points 100 m

                                         points high jump

                                           points 100 m

1,0000

,2792

p= ---

p=,085

                                         points high jump

,2792

1,0000

p=,085

p= ---


Because p (0,085)>0,05, we don´t reject null hypothesis, which is telling that correlation
coefficient is ZERO. It means that correlation coefficient is ZERO. Correlation is statistically
significant.