07 - 11 - 2014
Save $25 on orders of $100+, $50 on orders of $150+, and save $100 on orders of $250+ by entering code AFF2014 at store.elsevier.com!
Main menu
Orphus system

Parametric tests: Paired (dependent) t-test

Paired t-test is used to compare groups of dependent data, in which every point in one group corresponds the same point in another group. In our example (see Example 1) each row represents four dependent measurements which correspond one bacterial strain. By paired t-test we may compare inhibition zones around the disks with tea tree oil alone and tea tree oil with gatifloxacin, we may compare inhibition zones around thyme oil alone and thyme oil with gatifloxacin. Also we may compare inhibition zones around disk with tea tree oil and with thyme oil, etc.

Before performing statistical analysis let us split the file by variable “Species” in order to have separate results for E. coli and E. faecalis by clicking the Split button and specifying output by groups (see Splitting the file).

To specify paired-sample t-test:

1) Click the Analyze menu, point to Compare Means, and select Paired-Samples T-test… :

Paired-samples t-test

The Paired-Samples T-test dialog box opens:

Paired-samples t-test: specifying variables

2) Select the first variable from pair which will be compared (“TeaTree”); click the transfer arrow button . The selected variable is moved to the Paired Variable(s): list box.

3) Select the second variable from the pair which will be compared (“TeaTreeGati”); click the transfer arrow button . The selected variable is also moved to the Paired Variable(s): list box.

4) Select in the same way rest two variables (“Thyme”) and (“ThymeGati”) and also transfer them to the Paired Variable(s): list box:

Paired-samples t-test: specifying variables

5) Click the OK button. An Output Viewer window opens and displays the statistics.

The Output Viewer window contains three tables with statistical results for each species – paired samples statistics, paired samples correlations and paired samples test. In the Paired Samples Statistics table, mean, number of cases, standard deviation and standard error of mean are shown. The Paired Samples Correlations table contains number of cases, correlation coefficients and their significance. We can see that the data for inhibition zones of thyme oil alone and with gatifloxacin against E. coli there are correlated (correlation coefficient is -0.674, p = 0.016).

Among data for E. faecalis there are no correlations. The Paired Sample Test table contains results of t-test themselves. From these figures we can see that differences between inhibition zones around disks with tea tree oil alone and with gatifloxacin are not significant (p = 0.252 for E. coli and p = 0.145 for E. faecalis). In contrast, between inhibition zones for thyme alone and with gatifloxacin there are statistically significant differences against both tested species with p = 0.001.

Paired samples statistics: E. coli

Paired samples correlations: E. coli

Paired samples tests: E. coli

Paired samples statistics: E. faecalis

Correlations: E. faecalis

Paired samples tests: E. faecalis

blog comments powered by DISQUS back to top