It is easy to understand the meaning of relationship and effect (impact) in general. But when it comes to the statistical analysis it is a difficult conundrum, as one can determine the relationship between two variables using the correlation and effect can be analyzed using the regression. Is there further ways to differentiate?
Correlation does not imply causality, if X and Y are correlated it does not mean they have to functionally related to each other, it is possible that both have implicit dependence on a third variable Z.
Correlation is simply a relationship. Effect can be analysed by causation also (through cause and effect relationship)
Correlation does not imply causality, if X and Y are correlated it does not mean they have to functionally related to each other, it is possible that both have implicit dependence on a third variable Z.
Farheen Akram, I can appreciate your problem because variables can be related to each other in complex ways, including feedback loops, and certainly correlation does not imply causation.
Although it's not a perfect solution, perhaps you can propose the direction of causality from theoretical bases - and conduct your statistical analyses (e.g., multiple regression) on the basis of that.
Check how we did to discover relationships and impact:
The "effect" in Statistical Inferences and Analysis generally refers to the concept of a variable(s) having effect (fixed or random or pooled) on another (or a dependent variable [DV] taking on statistically different value in response to an intervention (observations) or treatment (experimental designs) or time(Time series)!
Effect Statistics are computed/estimated/infered. The two common usage are the Main Effect and Interaction Effect Statistics. These are the difference in either:
1. Difference in the means (variables/factors), i.e., the difference between the true population parameter and the null hypothesis value, also difference in population/group means for example that computed by "Cohen's d" in a post ANOVA or t-test.
2. the Correlation Coefficient [spearman's/product moment Correlation; Pearson's and Intra Class correlation or ICC.], and
3. Relative Frequency and Relative Risk Statistics [i.e., Odds].
*Note, each of these Effect Statistics comes in several varieties and with respect to studies and research questions.
#Stay safe
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression