At the inital stage, this might help. I am trying a simple explanation here, there could be said much more, but I think for beginners, this might be what you ware looking for?
One sample t-test: compare the mean of your sample to a given value (e.g. a group of students writes a test, you take the mean value of the points they achieved an compare it to a value given by the textbook which indicates "good performance". Question: does my sample value differ significantly from the value I compare it to?"
independent sample t-test: you compare the means of two groups, e.g. male and female students in the class. Question: do the mean values differ significantly?
paired sample t-test: you compare the same group (or closely related samples) at two different points in times. Example: students take a test at the beginning of the semester and at the end, you compare if there are significant improvements. Question: do the means differ significantly?
One way ANOVA: you have more than two groups (t-Tests are only for 1 group and an external value you want to compare to, or 2 groups (between or repeated)) and want to compare means. e.g. three classes of students and you want to compare how they did on a test.
Two way ANOVA: you have more than one category that determines your groups (more than one independent variable). e.g. you want to see the effect of being in different classes AND gender on the performance in a test. and how the factors might interact with each other. then you have two factors, thus a two way ANOVA. Questions: which class does better in the test? which gender does better? is a combination of both, class and gender, significant? (e.g. males in class A score high, females in class A score low, but males in class B score low and females in class B score high: that is an interaction of two factors showing that dependent on the class, gender effects perfomance differently).
At the inital stage, this might help. I am trying a simple explanation here, there could be said much more, but I think for beginners, this might be what you ware looking for?
One sample t-test: compare the mean of your sample to a given value (e.g. a group of students writes a test, you take the mean value of the points they achieved an compare it to a value given by the textbook which indicates "good performance". Question: does my sample value differ significantly from the value I compare it to?"
independent sample t-test: you compare the means of two groups, e.g. male and female students in the class. Question: do the mean values differ significantly?
paired sample t-test: you compare the same group (or closely related samples) at two different points in times. Example: students take a test at the beginning of the semester and at the end, you compare if there are significant improvements. Question: do the means differ significantly?
One way ANOVA: you have more than two groups (t-Tests are only for 1 group and an external value you want to compare to, or 2 groups (between or repeated)) and want to compare means. e.g. three classes of students and you want to compare how they did on a test.
Two way ANOVA: you have more than one category that determines your groups (more than one independent variable). e.g. you want to see the effect of being in different classes AND gender on the performance in a test. and how the factors might interact with each other. then you have two factors, thus a two way ANOVA. Questions: which class does better in the test? which gender does better? is a combination of both, class and gender, significant? (e.g. males in class A score high, females in class A score low, but males in class B score low and females in class B score high: that is an interaction of two factors showing that dependent on the class, gender effects perfomance differently).
If you're looking for SPSS help, see my Youtube Channel below. It has about 10 videos covering the basic tests in SPSS as well as some commands and actions in the program. I made these for my students, but others have found them helpful.
one-sample t-test: You want to compare mileage of cars, which one is performing better.
Independent sample t-test: You want to compare mileage of cars, based on only two different brands. Which one is performing better?
One-way ANOVA: You want to compare mileage of cars, based on only more than two different brands. Which one is performing better?
Paired sample t-test: You want to compare mileage of cars, you have implemented an extra part in the engine of the same brand car and wanted to study Before and After performance. Whether the part is better one or not?
Hi Shabana, welcome to the world of stats and analytics! :) Just imagine it this way:
What t-tests do (one-sample, independent, paired, etc) is to find out whether the sample or more than one sample belongs to the population. As you know there is a sample and a population and a sample is a representation of the population. So I can sample a group of my countrymen to study the whole population of my country, Singapore. The various sub-branches of the techniques such as one-sample, independent, paired, etc are just different ways to test it. Johanna definitely have spelled that out.
For ANOVA, it is a collection of tests that try to find out very precisely whether the groups you want to study come from the same population. It includes t-test, f-test, levene test, etc. But the most interesting test within ANOVA is the f-test - I would say. If you study the mathematical expression behind it, it has one of the most logical distribution in reality.
one-way ANOVA or two-way ANOVA or n-way ANOVA depends on the number of factors or independent categorical variables you have in the equation. If you have one categorical variable at the right side of the equation, then it is one-way ANOVA. They have significant impact in your research work as n-way ANOVA requires a study of interaction and main effects. I think you may want to leave the interaction and main effects to later exploration as it can get a little bit complex.