Repeated measures - missing values

Moderators: statman, Analyst Techy, andris, Fierce, GerineL, Smash

gracedooley
Posts: 1
Joined: Tue Oct 13, 2015 9:14 am

Repeated measures - missing values

Postby gracedooley » Tue Oct 13, 2015 9:29 am

Hello!

Bear with me while I do my best to explain my problem. Briefly; I am testing whether acclimating plants to low levels of salt can create a tolerance in their offspring, over 3 generations. However, in my first generation the plants died before setting seed. See below:

Gen 1: high level of salt, died without setting seed.
Gen 2: medium level of salt, survived and set seed.
Gen 3: identical salt level to Gen 1; testing to see whether an acclimated parent plant could provide a tolerance.

In each generation there are 128 plants overall with 4 treatments (32 in each treatment). However, for half of the first generation and most of the third generation I have masses of missing data where the plants either died or failed to germinate.

I have height, weight, flower number and 5 points of chlorophyll fluorescence as dependent variables, with time and salt amount as independent. With 12 data points in each generation, repeated measures one-way ANOVA seems to be the best option but with so much data missing, would this not affect the output?

Thanks for your time.
GerineL
Moderator
Posts: 1477
Joined: Tue Jun 10, 2008 4:50 pm

Re: Repeated measures - missing values

Postby GerineL » Mon Nov 09, 2015 10:54 am

it it an option (theoretically) to leave out this first category? Or can you set the outcome to 0 or something?
This is absolutely not my field so I am not sure, but this soulds like a problem more biologists might run into, so maybe check the literature?

Who is online

Users browsing this forum: No registered users and 2 guests

cron