Variance analysis and t-test are both frequently used methods in studying and analyzing data, and they are both studying a difference relationship. Let me briefly talk about these two analysis methods. What is analysis of variance? What is a t-test?
1, analysis of variance
Analysis of variance, also known as "F test", is used to test the significance of the average difference between two or more samples. The basic idea of variance analysis is to determine the influence of controllable factors on the research results by analyzing the contribution of variation from different sources to the total variation.
According to the different independent variable x in the study, ANOVA can be subdivided. When the number of x is 1, we call it one-way variance; When x is 2, it is a two-factor variance; When x is 3, it is called three-factor variance, and so on. When x exceeds 1, it is collectively called multivariate variance.
In this paper, SPSS sau-online SPSS analysis software is used as a tool to introduce in detail.
2. Classification of analysis of variance
One-way ANOVA: used to analyze the relationship between classified data and quantitative data. When using one-way ANOVA, the sample size of each option needs to be greater than 30, for example, the sample sizes of men and women are 100 and 120 respectively. If the sample size of an option is too small, the group should be merged first. For example, when studying the different attitudes of samples of different age groups to the research variables, there are only 20 samples younger than 20 years old, so it is necessary to compare the options younger than 20 years old with another group (such as 20 ~).
If the options cannot be combined, for example, to study the differences in attitudes of different professional samples to variables, the majors of the research samples are divided into four majors: marketing, psychology, education and management. These four majors are independent of each other and cannot be merged, but the sample size of marketing major is only 20, which is not representative. Therefore, we can consider screening marketing majors first, that is, only comparative psychology. When there are more than three groups in the three majors of education and management, and the differences are significant, we can consider using backtesting to further compare the differences between specific groups.
Please click to enter a picture description.
Two-factor analysis of variance: used to analyze the relationship between classified data (2) and quantitative data, such as the difference of online shopping satisfaction between gender and education level of researchers; And whether there are differences in online shopping satisfaction between men and women with different academic qualifications; Or whether there are differences in online shopping satisfaction between different sexes with the same academic qualifications.
Please click to enter a picture description.
Multivariate analysis of variance: for example, researchers test whether a new drug is effective on cholesterol levels; The researcher * * * recruited 72 subjects, 36 males and 36 females respectively, and made a breakdown of the use of new drugs and common drugs by males and females respectively; At the same time, patients with hypertension may interfere with new drugs, so researchers will also consider whether the subjects have hypertension. So in the end, X*** is divided into three parts, namely, drugs (old drugs and new drugs), gender, and whether they have hypertension; Y is the cholesterol level. Therefore, it is necessary to carry out three-factor variance analysis, that is, multi-factor variance analysis.
Please click to enter a picture description.
3, t test
T test, mainly used for small sample size (such as n
4, t test classification
T-test * * * is divided into three methods, namely independent sample T-test, paired sample T-test and single sample T-test.
Independent sample t-test: Independent sample t-test compares the differences between two groups of options, such as men and women. Relatively speaking, independent sample T-test is used more frequently in experimental comparison, especially in biological and medical related fields. In view of the questionnaire survey.
Independent sample t-test and paired sample t-test are different in function, and both compare the differences between the two groups. However, there are substantial differences between them. If we compare the differences of different gender and marital status (married and unmarried) samples on a certain variable, we should use independent sample t test. If there is a pairing relationship between the comparison groups, only the paired sample t test can be used, and the pairing relationship refers to the similar relationship between the experimental group and the control group. In addition, the number of samples in independent sample t-test can be unequal, while the number of samples in paired sample t-test needs to be completely equal.
5, when to use t test? When to use ANOVA?
The difference between variance and t-test is that the independent variable x of t-test can only be divided into two categories, such as male and female. If x is in three categories, such as below undergraduate course and above undergraduate course; Only analysis of variance can be used at this time.
In the choice of methods, questionnaire research usually adopts variance analysis, but some majors, such as psychology, pedagogy or normal universities, often use t-test for analysis when conducting experimental research. In addition, there are many differences between ANOVA and T-test, and only one of them can be used in some analyses.