Mean differences do not guarantee co-variances of variables.

[Perhaps an example will make this clear:

If you compare the average temperature, number of spelling mistakes at school, number of carnival parades and number of words spoken in 50 German cities in February and July respectively, you will find significant differences four times. (The carnival month of February is here in winter. July is a vacation month, and is three days longer, so more words are spoken overall). Would you have expected temperature, spelling errors at school, and carnival parades to explain how often people speak to each other?]

This means that all variables have changed independently of each other over time. Thus, self-efficacy and performance may have increased on average, but without the learners whose self-efficacy has increased also having an increase in learning.

Depending on the number of students participating, I recommend listing and comparing the variables or doing a cluster analysis to find groups (e.g. increased self-efficacy and increased performance, increased self-efficacy and unchanged performance, etc.).

Good afternoon dear Randel Estacio

You should have first measured the relationship between those variables before determining the influence.

1.-You only measured the relationship between the variables: (cognitive load, knowledge gain, and self-efficacy)

2.-You did not measure the correlation strength between those variables

3.-You should have measured the strength of the correlation between the variables: (cognitive load, knowledge gain, and self-efficacy)

4.-Only based on the third point, only there, could you determine the influence of the graphic organizers on these variables.

Remember that a relationship is not enough to prove a consequence.