When you are an independent and dependent variable , it is best to use regression. Linear regression is the easiest and best solution.
Use the automatic linear modeling SPSS option. It will have interesting shapes.
You have a very simple job.
Hi Aaron
Some kind of regression analysis is suitable. However, what one depends on the distribution of the dependent variable. For example, if it is dichotomous (only has 2 values), then you would undertake logistic regression. Whichever type of regression you do, the usual null hypothesis is that the regression coefficient is zero.
La Trobe will have statistical consultants available. You could approach one of them, or alternatively, look up on Youtube on how to do regression analysis.
Statistics cannot prove causation. To prove that you need an appropriate experiment design.
I wish to state that hypothesis test cannot determine causality between dependent and independent variables. Hypothesis test is based on probability..... Estimation of relationship or association between variables based on some degree of confidence or error.
The hypothesis test and associated analysis depends on the type of variable... Nominal, ordinal or numerical.
To make it simple there are two broad categories of tests. One is comparision. if u are comaring two variables you will use t/z test. if comparing more than two u will use ANOVA. Second category is effects. if you have one independent and one depend variable you will use simple regression. in case you have one depedent variable and more than one independent variables you will use multiple regression. ANCOVA and other related test i have not discussed to keep the answer simple.
depend on the type of independent and dependent variables(Measurable or categorical). If both are measurable then use simple linear regression. If dependent is categorical the use either binary logistic or multi nominal logistic regression
Hi
1. You should study your variables and your research questions.
2. The research questions have to be transformed to hypotheses.
3. Then on the bases of your data and the assumptions of statistical measures you can develop the suitable tests.
Regards,
Zuhair
Hi,
If you want to study the effect of independent variable on dependent variable i advise you to use simple regression model.
Regards
For 2 independent samples, we use T-test (SPSS) : Analyze > Compare means > Independent samples T-test.
But in your case, best method is
(Regression)
(SPSS) : Analyze > Regression > Linear... on the top menu. Then, Transfer the independent variable into the (Independent(s): ) box and the dependent variable into the (Dependent: ) box.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS