+ Reply to Thread
Results 1 to 4 of 4

Thread: re: basic question on why my predictor is not significant?

  1. #1
    Points: 7, Level: 1
    Level completed: 13%, Points required for next Level: 43

    Posts
    2
    Thanks
    0
    Thanked 0 Times in 0 Posts

    re: basic question on why my predictor is not significant?




    I am puzzle by why my predictor was not significant so I would like to seek your expertise. My regression model does not provide significant on the predictor value. Would you suggest that because I have too many levels on my predictor value? I have 500 levels on my predictor values meaning predictor contains code of "A1" through "A500". My model just have two variables. One outcome variable naming y, and one predictor value naming x. Outcome variable is alternate between 0 and 1, and the predictor value is contain all values of A1 through A500. Furthermore , I have 1000 observations in my data. My question is that why my predictor variable is not significant when I ran the glm analysis? My guess is that because my predictor contains so many levels meaning 500 levels. Is this the problem why it would not be significant? Please help, and also if I try to reduce the number of levels of my predictor, then would it be significant? Thanks for all of your help in advance.

    regression

  2. #2
    Omega Contributor
    Points: 38,289, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    6,992
    Thanks
    397
    Thanked 1,185 Times in 1,146 Posts

    Re: basic question on why my predictor is not significant?

    What program are you using? Are you getting any warning messages? Is it running at all? Can you run the model with fewer levels to see if it will run? Can the levels be turned into a continous values? Lots of questions, but 500 levels seems like quite a few.
    Stop cowardice, ban guns!

  3. #3
    Points: 7, Level: 1
    Level completed: 13%, Points required for next Level: 43

    Posts
    2
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: basic question on why my predictor is not significant?

    Hello, I am using R, and none of my predictor value was significant. No, I can not turn the predictor level to continous.
    Here is my sample output:
    > summary (reg)

    Call:
    glm(formula = x ~ y, family = binomial("logit"), data = df)

    Deviance Residuals:
    Min 1Q Median 3Q Max
    -1.66511 -0.75853 -0.00013 0.86325 1.66511

    Coefficients:
    Estimate Std. Error z value Pr(>|z|)
    (Intercept) -1.857e+01 3.261e+03 -0.006 0.995
    yA100 1.747e+01 3.261e+03 0.005 0.996
    yA101 1.857e+01 3.261e+03 0.006 0.995
    yA102 1.747e+01 3.261e+03 0.005 0.996
    yA103 1.857e+01 3.261e+03 0.006 0.995
    yA104 1.747e+01 3.261e+03 0.005 0.996
    yA105 1.966e+01 3.261e+03 0.006 0.995
    yA106 2.271e-07 4.612e+03 0.000 1.000
    yA107 1.857e+01 3.261e+03 0.006 0.995
    yA108 1.747e+01 3.261e+03 0.005 0.996
    yA109 1.857e+01 3.261e+03 0.006 0.995
    yA11 1.857e+01 3.261e+03 0.006 0.995
    yA110 1.747e+01 3.261e+03 0.005 0.996
    yA111 1.747e+01 3.261e+03 0.005 0.996
    yA112 1.747e+01 3.261e+03 0.005 0.996
    yA113 1.857e+01 3.261e+03 0.006 0.995
    yA114 1.747e+01 3.261e+03 0.005 0.996
    yA115 1.747e+01 3.261e+03 0.005 0.996
    yA116 1.966e+01 3.261e+03 0.006 0.995
    yA117 2.267e-07 4.612e+03 0.000 1.000
    yA118 1.747e+01 3.261e+03 0.005 0.996
    yA119 2.273e-07 4.612e+03 0.000 1.000
    yA12 1.857e+01 3.261e+03 0.006 0.995
    yA120 1.966e+01 3.261e+03 0.006 0.995
    yA121 2.273e-07 4.612e+03 0.000 1.000
    yA122 1.747e+01 3.261e+03 0.005 0.996
    yA123 1.747e+01 3.261e+03 0.005 0.996
    yA124 1.747e+01 3.261e+03 0.005 0.996
    yA125 2.262e-07 4.612e+03 0.000 1.000
    yA126 1.966e+01 3.261e+03 0.006 0.995
    yA127 1.747e+01 3.261e+03 0.005 0.996
    yA128 2.280e-07 4.612e+03 0.000 1.000
    yA129 1.857e+01 3.261e+03 0.006 0.995
    yA13 1.857e+01 3.261e+03 0.006 0.995
    yA130 1.857e+01 3.261e+03 0.006 0.995
    yA131 1.747e+01 3.261e+03 0.005 0.996
    yA132 2.267e-07 4.612e+03 0.000 1.000
    yA133 1.966e+01 3.261e+03 0.006 0.995
    yA134 2.265e-07 4.612e+03 0.000 1.000
    yA135 1.747e+01 3.261e+03 0.005 0.996
    yA136 1.857e+01 3.261e+03 0.006 0.995
    yA137 1.747e+01 3.261e+03 0.005 0.996
    yA138 2.267e-07 4.612e+03 0.000 1.000
    yA139 1.857e+01 3.261e+03 0.006 0.995
    yA14 2.277e-07 4.612e+03 0.000 1.000
    yA140 2.273e-07 4.612e+03 0.000 1.000
    yA141 1.747e+01 3.261e+03 0.005 0.996
    yA142 1.747e+01 3.261e+03 0.005 0.996
    yA143 2.252e-07 4.612e+03 0.000 1.000
    yA144 1.747e+01 3.261e+03 0.005 0.996
    yA145 1.747e+01 3.261e+03 0.005 0.996
    yA146 2.262e-07 4.612e+03 0.000 1.000
    yA147 1.747e+01 3.261e+03 0.005 0.996
    yA148 2.267e-07 4.612e+03 0.000 1.000
    yA149 1.747e+01 3.261e+03 0.005 0.996
    yA15 1.857e+01 3.261e+03 0.006 0.995
    yA150 1.857e+01 3.261e+03 0.006 0.995
    yA151 1.857e+01 3.261e+03 0.006 0.995
    yA152 1.747e+01 3.261e+03 0.005 0.996
    yA153 1.747e+01 3.261e+03 0.005 0.996
    yA154 1.857e+01 3.261e+03 0.006 0.995
    yA155 1.747e+01 3.261e+03 0.005 0.996
    yA156 2.255e-07 4.612e+03 0.000 1.000
    yA157 1.857e+01 3.261e+03 0.006 0.995
    yA158 1.747e+01 3.261e+03 0.005 0.996
    yA159 2.262e-07 4.612e+03 0.000 1.000
    yA16 2.273e-07 4.612e+03 0.000 1.000
    yA160 2.267e-07 4.612e+03 0.000 1.000
    yA161 1.747e+01 3.261e+03 0.005 0.996
    yA162 2.272e-07 4.612e+03 0.000 1.000
    yA163 1.857e+01 3.261e+03 0.006 0.995
    yA164 1.747e+01 3.261e+03 0.005 0.996
    yA165 2.272e-07 4.612e+03 0.000 1.000

  4. #4
    Points: 4,664, Level: 43
    Level completed: 57%, Points required for next Level: 86
    kiton's Avatar
    Location
    Corn field
    Posts
    234
    Thanks
    47
    Thanked 51 Times in 46 Posts

    Re: basic question on why my predictor is not significant?


    First of all, as far as I understood, your DV (y) is binary, that is has values of either 0 or 1. In this case the proper type of analysis would be logistic regression. Secondly, could you please post the descriptive statistics for your IV (mean, SD, min, max).

+ Reply to Thread

           




Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts






Advertise on Talk Stats