0
$\begingroup$

I have a question related to application of a linear mixed effect model.

I have land use data in percentage, which is the predictor, and a water quality score (e.g. 100) as response variable for 100 areas. However, for each area I have only one set of data both for predictors and response variable. I was wondering, can I use a linear mixed effect model to determine which land use is significantly contributing the water quality score? This question arises as I don't have repeated measures for each area. I am providing some of my data to better explain my question.

Data:

Ag.land  Forest  Urban_H  Urban_M   Urban_L  Wetland        Water quality score
Area-1    4.9      65      1.09      .18       .40       19.09           73.7
Area-2     6.73     19.72   24.28    0.36      1.91     27.61            57.03

Similarly, I have data for Area-3 up to Area-100.

$\endgroup$
2
  • 2
    $\begingroup$ Is your intention to learn about linear mixed models or to understand the relationships between water quality scores and land use? If it's the latter, then I recommend that you completely rewrite this post: drop all references to mixed models and ask instead about how you might create a useful model. $\endgroup$
    – whuber
    May 1, 2015 at 14:23
  • $\begingroup$ Whuber, Thank you very much! I want to understand the relationship between water quality score and land use matrices. $\endgroup$
    – Raj
    May 4, 2015 at 12:22

1 Answer 1

1
$\begingroup$

If you don't have repeated measures or spatial clustering then you don't need a mixed model. You can just use some (non-mixed) form of regression.

Of course, you could pose this as a mixed model with no random effect, but that wouldn't really be a mixed model. (That is, you could use e.g PROC MIXED in SAS to do a regular regression).

$\endgroup$
0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.