Re: [ECOLOG-L] statistical advice on quantile regressions etc.
Mikaela - The package quantreg in R is a great resource for conducting quantile regression. Also, Brian Cade with USGS in Ft. Collins, CO has published some papers dealing with quantile regression of ecological data. I've linked to his USGS page below. You can scroll through the publications listed there to find some good resources: http://www.fort.usgs.gov/staff/staffprofile.asp?StaffID=115 If you need additional help, let me know. Steven On Wed, 13 Feb 2013 15:31:27 -0600, Mikaela Gioia Howie mikaelaho...@hotmail.com wrote: Hi eco-loggers! I am attempting to publish my master's thesis work for what feels to be the nth time and I am looking for some statistical advice or someone who would be willing to coach me through some statistical queries. Specifically, I am considering running quantile regressions but I have not used these in the past and I am a little unsure if they are the correct statistical procedure to use with my data. I currently use SPSS and I have begun to make the transition to R, so someone who has expertise in either would be very helpful! Any advice is welcome! Thanks! Mikaela Mikaela G. Howie
[ECOLOG-L] statistical advice on quantile regressions etc.
Hi eco-loggers! I am attempting to publish my master's thesis work for what feels to be the nth time and I am looking for some statistical advice or someone who would be willing to coach me through some statistical queries. Specifically, I am considering running quantile regressions but I have not used these in the past and I am a little unsure if they are the correct statistical procedure to use with my data. I currently use SPSS and I have begun to make the transition to R, so someone who has expertise in either would be very helpful! Any advice is welcome! Thanks! Mikaela Mikaela G. Howie
Re: [ECOLOG-L] Statistical advice
Hi Bill: Without having more information about the design (specific hypothesis) and given the temporal replication, you might wish to consider using occupancy models (http://www.uvm.edu/envnr/vtcfwru/spreadsheets/occupancy/occupancy.htm) rather than repeated measures. These types of models can also be applied for interactions between species (see Chapter 8; Occupancy Estimation and Modeling. MacKenzie et. al. 2006. Academic Press). Brian Campbell Date: Mon, 30 Jun 2008 13:43:00 -0500 From: [EMAIL PROTECTED] Subject: [ECOLOG-L] Statistical advice To: ECOLOG-L@LISTSERV.UMD.EDU Good afternoon all: I am looking for some statistical advice, in a situation that has me temporarily stumped. We have data which includes a categorical predictor variable (a landscape attribute, habitat patch size), two continuous dependent variables (measures of plant and rodent abundance), and many years of observations. Experimental hypotheses involve the question of how patch size affects organism abundance, and also about correlations between plant and rodent abundance. This seems to be set up exactly for the repeated measures ANOVA function in SPSS within the GLM section, only no information is given in the printout about associations between the dependent variables. What would you recommend we do to formally investigate the relations between plant and rodent abundance (the dependent variables), in the light of time and patch size? So far we can run a RMANOVA to investigate time and patch size, and then to run separate analyses (e.g. correlations within each year) to look at the association between plant and rodent abundance, but there may be a more holistic way to do this. Thanks for any advice you can give. Bill Cook William M. Cook Assistant Professor Department of Biological Sciences St. Cloud State University 720 4th Avenue South St. Cloud, MN 56301 USA Phone: (320) 308-2019 E-mail: [EMAIL PROTECTED] _ Send a smile, make someone laugh, have some fun! Start now! http://www.freemessengeremoticons.ca/?icid=EMENCA122
[ECOLOG-L] Statistical advice
Good afternoon all: I am looking for some statistical advice, in a situation that has me temporarily stumped. We have data which includes a categorical predictor variable (a landscape attribute, habitat patch size), two continuous dependent variables (measures of plant and rodent abundance), and many years of observations. Experimental hypotheses involve the question of how patch size affects organism abundance, and also about correlations between plant and rodent abundance. This seems to be set up exactly for the repeated measures ANOVA function in SPSS within the GLM section, only no information is given in the printout about associations between the dependent variables. What would you recommend we do to formally investigate the relations between plant and rodent abundance (the dependent variables), in the light of time and patch size? So far we can run a RMANOVA to investigate time and patch size, and then to run separate analyses (e.g. correlations within each year) to look at the association between plant and rodent abundance, but there may be a more holistic way to do this. Thanks for any advice you can give. Bill Cook William M. Cook Assistant Professor Department of Biological Sciences St. Cloud State University 720 4th Avenue South St. Cloud, MN 56301 USA Phone: (320) 308-2019 E-mail: [EMAIL PROTECTED]
