I am doing PLS-SEM.
According to Joseph F. Hair and his colleagues, the acceptable level of R2 depends on the research context with higher values indicate higher level of predictive accuracy. In some disciplines such as consumer behaviour, values of 0.20 are considered high while in scholarly research that focuses on marketing issues, 0.75, 0.50 and 0.25 are described substantial, moderate and weak, respectively.
I cannot find such a description in educational research.
Is there any description of acceptable R2 value for structural models in educational research?
In social science research, the value of R-Square matters when you are having some complex relationships in Model. R-Square is a base to check the change effect in moderation based model. Change in R-Square is a prove of moderation. However, The perfect value of R-Square is still controversy.
Myo,
Be careful using the word prediction. R Squared indicates the amount a shared variation between two or more variables. The R Squared value is not an indication of causality; just co-variance. There are some situations when using SEM where a case for causality can be made (see D Prime).
The importance of a large R Squared value depends on the power of the test which depends on the size of the sample and the degrees of freedom. What significance level are you prepared to accept? Even a large R value is meaningless unless it reaches the required level of significance. Significance means the probability that your could have gotten an R value this large just by chance. Also, there are many other statistical issues to consider when using SEM or regression besides the significance level. BTW I don't like the use of the word significance; reliability level is more apt as significance implies importance which is a wholly difference construct.
If you have supervisors for this research, I suggest you discuss the R Squared question with them. Alternately, look at publications in this area and see what level of R Squared they regard as meaningful.
George
Is there any acceptable R2 value for educational research in PLS-SEM?
You can also refer to the following RG links:
https://www.researchgate.net/post/what_is_the_acceptable_r-squared_value
https://www.researchgate.net/post/what_is_a_good_r_square_value_in_regression_analysis
I'm not sure, but i presume that the answer of Naveed Ahmad is quite reasonable
CColaço
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology