I know that the independent sample T test proves that the experimental group and the control group are at the same level. Then the paired sample t-test proves that there is a significant difference (progress) between the back and the front. There is also a Lickert scale questionnaire to prove the change of attitude.

You mean that you should use t-test to analyze this scale. First of all, t-test is suitable for numerical variable data, and it also requires the normal distribution and variance homogeneity of two independent samples.

This scale collects the frequency of each grade, among which 1 2 3 4 5 is meaningless, just representing different grades.

If this scale collects the data of the experimental class and the ordinary class, you should compare the differences between them. If the chi-square test conditions are met, chi-square can be used. If not, other testing methods can be considered as the case may be.

The analysis-descriptive statistics-crosstab in spss can be tested by chi-square.