Notices
Results 1 to 7 of 7

Thread: Urgent statistics help

  1. #1 Urgent statistics help 
    Forum Freshman
    Join Date
    Nov 2011
    Posts
    5
    Hi there I need some help! What statistical tests should i use/consider (pref in SPSS) in regards to an invertebrate survey i have carried out comparing urban street trees to urban park trees. I am looking at species richness and abundance between the two sites and also other variables such as tree species, tree width at base, distance from road/tarmac, tree base substrate and weather at time of collection. any help would be greatly appreciated!


    Reply With Quote  
     

  2.  
     

  3. #2  
    Forum Professor wallaby's Avatar
    Join Date
    Jul 2005
    Location
    Australia
    Posts
    1,521
    If you want to know only if there is a difference between the species richness/abundance, tree width, etc. Between the two sites then what you want to do is perform a one way ANOVA on each data set, ie. species richness vs site, abundance vs site, tree width vs site. This test will tell you if the mean richness/abundance or width are significantly different between the sites.

    ANOVA is used when the response variable is continuous and is dependent on a categorical variable. If however you wanted to test a continuous response to one (or more) continuous variable then you need to fit a linear model, i don't know about SPSS but some statistics programs also run the appropriate tests that determine if a variable is significant when fitting a linear model. (the actual tests for significance involve F-tests and a headache)


    Reply With Quote  
     

  4. #3  
    Forum Freshman
    Join Date
    Nov 2011
    Posts
    5
    So the suggestion here would be in using the ANOVA test use it to compare say urban Species richness/abundance against width, tree species etc then do the same for park species abundance/richness to see if any of the variables are having an effect upon the spp richness/abundance. Then to compare the species richness/abundance of the parkland v.s the species richness/abundance of the urban trees i merely do the one way ANOVA comparing the spp rich urban v.s parkland then the same for abundance?
    Reply With Quote  
     

  5. #4  
    Forum Freshman
    Join Date
    Nov 2011
    Posts
    6
    Statistics is one of the toughest concepts of mathematics and we try best to make it simple for you. I didn't clear these topic. will you please help me in the areas listed below:
    • Mean
    • Median
    • Mode
    • Range
    • Probability

    My examinations dates are near so i afraid of statistics so please provide me some
    statistics help
    when you will reply i will keep post my questions and my doubts regarding statistics..
    Reply With Quote  
     

  6. #5  
    Forum Professor wallaby's Avatar
    Join Date
    Jul 2005
    Location
    Australia
    Posts
    1,521
    Quote Originally Posted by sylver View Post
    So the suggestion here would be in using the ANOVA test use it to compare say urban Species richness/abundance against width, tree species etc then do the same for park species abundance/richness to see if any of the variables are having an effect upon the spp richness/abundance. Then to compare the species richness/abundance of the parkland v.s the species richness/abundance of the urban trees i merely do the one way ANOVA comparing the spp rich urban v.s parkland then the same for abundance?
    If you want to know if being a street tree or an urban park tree affects your species abundance you use the ANOVA test. (think of species abundance as your dependent variable and the site as being the independent variable) The ANOVA test will tell you if the mean abundance is significantly different between the locations. You can compare any of your continuous variables to the site at which the measurements were taken using this test, so tree width vs site but probably not tree type vs site. (i just realised i'm assuming species richness/abundance are continuous variables?)

    If you want to compare two, or more, continuous variables you need to formulate a linear model, this is a little more involved but i'm sure the statistics package will do it for you and tell you if a particular variable is of significant influence on the output variable. (this is probably best illustrated by example so if this is what you're interested in doing then just give an example) Comparing a continuous variable with multiple continuous and categorical variables can also be done, but again it is a little more involved.

    Sorry if this is really vague, there's just a large body of theory behind what tests to use and when.

    Quote Originally Posted by geeknick
    Statistics is one of the toughest concepts of mathematics and we try best to make it simple for you. I didn't clear these topic. will you please help me in the areas listed below:
    • Mean
    • Median
    • Mode
    • Range
    • Probability

    My examinations dates are near so i afraid of statistics so please provide me some statistics help when you will reply i will keep post my questions and my doubts regarding statistics..


    Welcome to the forum geeknick. From the topics you listed i'd say these videos will be of interest to you, the explanations are quite clear and simple. If you have any further questions though you might want to post them in a new thread within the mathematics forum.
    Last edited by wallaby; November 5th, 2011 at 04:22 AM. Reason: added to post
    Reply With Quote  
     

  7. #6  
    Forum Freshman
    Join Date
    Nov 2011
    Posts
    5
    Quote Originally Posted by wallaby View Post
    Quote Originally Posted by sylver View Post
    So the suggestion here would be in using the ANOVA test use it to compare say urban Species richness/abundance against width, tree species etc then do the same for park species abundance/richness to see if any of the variables are having an effect upon the spp richness/abundance. Then to compare the species richness/abundance of the parkland v.s the species richness/abundance of the urban trees i merely do the one way ANOVA comparing the spp rich urban v.s parkland then the same for abundance?
    If you want to know if being a street tree or an urban park tree affects your species abundance you use the ANOVA test. (think of species abundance as your dependent variable and the site as being the independent variable) The ANOVA test will tell you if the mean abundance is significantly different between the locations. You can compare any of your continuous variables to the site at which the measurements were taken using this test, so tree width vs site but probably not tree type vs site. (i just realised i'm assuming species richness/abundance are continuous variables?)

    If you want to compare two, or more, continuous variables you need to formulate a linear model, this is a little more involved but i'm sure the statistics package will do it for you and tell you if a particular variable is of significant influence on the output variable. (this is probably best illustrated by example so if this is what you're interested in doing then just give an example) Comparing a continuous variable with multiple continuous and categorical variables can also be done, but again it is a little more involved.

    Sorry if this is really vague, there's just a large body of theory behind what tests to use and when.

    Quote Originally Posted by geeknick
    Statistics is one of the toughest concepts of mathematics and we try best to make it simple for you. I didn't clear these topic. will you please help me in the areas listed below:
    • Mean
    • Median
    • Mode
    • Range
    • Probability

    My examinations dates are near so i afraid of statistics so please provide me some statistics help when you will reply i will keep post my questions and my doubts regarding statistics..


    Welcome to the forum geeknick. From the topics you listed i'd say these videos will be of interest to you, the explanations are quite clear and simple. If you have any further questions though you might want to post them in a new thread within the mathematics forum.
    OK well I have the ANOVA down in my mind now and that will indeed be the main bit of stats used however I am interested in also doing the Linear Model http://www.cchil.org/cru/images/educ...293ed4c8da.pdf that is what SPSS says about it so if you could just give me a pointer with that information as to what means what as this is my greatest weakness I can look up statistical tests all day long but knowing what data set is what in the test is always my downfall. If by continious variable you mean it has to have a decimal element to it the species richness does not it is merely the number of species in the sample site? For example you can't get 4.3 species? Unless I am misunderstanding you which is highly probable! If you could give me an example with the variables i have mentioned for the linear model it would be greatly appreciated.
    Reply With Quote  
     

  8. #7  
    Forum Professor wallaby's Avatar
    Join Date
    Jul 2005
    Location
    Australia
    Posts
    1,521
    Quote Originally Posted by sylver View Post
    OK well I have the ANOVA down in my mind now and that will indeed be the main bit of stats used however I am interested in also doing the Linear Model http://www.cchil.org/cru/images/educ...293ed4c8da.pdf that is what SPSS says about it so if you could just give me a pointer with that information as to what means what as this is my greatest weakness I can look up statistical tests all day long but knowing what data set is what in the test is always my downfall. If by continious variable you mean it has to have a decimal element to it the species richness does not it is merely the number of species in the sample site? For example you can't get 4.3 species? Unless I am misunderstanding you which is highly probable! If you could give me an example with the variables i have mentioned for the linear model it would be greatly appreciated.
    A continuous variable represents a measurement that can take on any value, so in the strictest sense species abundance is a discrete variable. However if your data range takes on a large range of values compared to the independent variable then you can still use tests like ANOVA and linear regression, you just have to be mindful of a slight error and you need to check the assumptions of such tests. (The link you provided gives the criteria for this) The ANOVA between species abundance and site should still be valid though, unless the data range for species abundance is very limited, since the variable 'Site' takes on two values.

    Just follow the instructions in the link you provided. Try using species abundance as your outcome variable, site as the catagorical variable and tree width as the continuous covariate. Then once you get a result follow the "assumptions testing" instructions and see what SPSS has to say.

    As an example of when a linear model would be appropriate, if the tree width data had a range of 1-10 meters and species abundance had a range of 1-100 then you could use linear regression to generate the model that relates the species abundance to the tree width. i.e. get an equation of the form, Species Abundance = A * Tree Width + B. In this case the species abundance could be treated as approximately continuous.
    Reply With Quote  
     

Similar Threads

  1. Statistics notation
    By ampakine in forum Mathematics
    Replies: 4
    Last Post: April 30th, 2011, 06:41 PM
  2. UNREPORTED CRIMES STATISTICS
    By WVBIG in forum Pseudoscience
    Replies: 3
    Last Post: June 22nd, 2008, 07:04 PM
  3. Statistics query
    By schizometric in forum Personal Theories & Alternative Ideas
    Replies: 3
    Last Post: February 11th, 2008, 05:50 AM
  4. art and statistics
    By Feifer in forum Art and Culture
    Replies: 4
    Last Post: January 6th, 2008, 05:59 PM
  5. the value and distortion of statistics
    By Maxime Maison in forum Philosophy
    Replies: 9
    Last Post: January 3rd, 2006, 10:14 AM
Bookmarks
Bookmarks
Posting Permissions
  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •