Need help running a mixed effect model with appropriate columns

[#1]

Hello community, 

I am interested in running a mixed effect model on on the influences of sex, location, and mass on the independent variable (y) mercury concentration. Currently, I have my columns and there properties as listed:

1: Watershed name

data type : character

model type: nominal

2: Adjusted Clawless Mass or (mass)'

data type: numeric

modeling type: continuous

3: sex/gravid

data type: numeric

model type: ordinal

4: mercury concentration

data type: numeric

model type: continuous

Currently, I was not able to find anything on what properties the columns need to be set as whether the fixed effects need to be numeric or character and continuous or ordinal or nominal. Same goes for my random effect (that being the watershed column) needing to be numeric or character and continuous, or ordinal, or nominal.

The effects I am trying to run are as followed

Fixed effects : sex and adjusted clawless mad

Random effects : Watershed name

Output or independent variable (y) : mercury concentration.

Do any of you professionals have any input on a way to set the column properties up to give me an accurate output and significance for what I am trying to test? any and all recommendations would be useful.

[#2]

Hi,

The way that you have defined the data and modelling types for your variables looks good to me.

Except I am not sure why sex/gravid would be an ordinal modelling type. Ordinal modelling type is for categorial variables that have an order. An example would be portion size of fries - small, medium, or large. It is categorical because it can only be small, medium, or large. But there is an order: large > medium > small. Another example would be where you have a variable for age divided into ranges, e.g. 0-18, 19-59, 60+.

Regards,

Phil