Fdtxjr

Does capillary rise violate hydrostatic paradox?

Can you take a "free object interaction" while incapacitated?

Why does the frost depth increase when the surface temperature warms up?

Started in 1987 vs. Starting in 1987

How do I lift the insulation blower into the attic?

Trouble reading roman numeral notation with flats

Friend wants my recommendation but I don't want to give it to him

Weird lines in Microsoft Word

Why does a 97 / 92 key piano exist by Bosendorfer?

Mortal danger in mid-grade literature

How to split IPA spelling into syllables

Why is "la Gestapo" feminine?

Is there a POSIX way to shutdown a UNIX machine?

What should be the ideal length of sentences in a blog post for ease of reading?

Would a primitive species be able to learn English from reading books alone?

Put the phone down / Put down the phone

Should I warn a new PhD Student?

How would a solely written language work mechanically

categorizing a variable turns it from insignificant to significant

Do native speakers use "ultima" and "proxima" frequently in spoken English?

Travelling in US for more than 90 days

How do you say "Trust your struggle." in French?

Reasons for having MCU pin-states default to pull-up/down out of reset

A seasonal riddle

GLM model not corresponding to exploratory analysis (R-studio)

Quantitative Analysis of Structure of Gaussian Mixture ModelTime series analysis ARCH($m$) modelTopics under Model Based Cluster AnalysisTime series analysis, moving-average model, ARMA modelsensitivity analysis of a modelFind the MLE of a GLMConstructing a 95% Confidence Interval (using output from r-studio)Why use the log_e function as link function for Poisson data in a GLM modelAnalysis of model fitDefinition of the fitted values of a GLM?

0

$begingroup$

I am working with a dataset from a job center that specialises in finding work placements for individuals with a criminal record.

There are 2 continuous factors (age, number of pre-convictions) and 4 discrete factors (placement, re-convicted, age group (grouped into 3 levels) and number of pre-convictions (grouped into 3 levels).

My exploratory analysis (consisting of Pearson's Chi-squared test, confidence regions, box plots and barcharts) suggested a strong relationship between number of pre-convictions and getting re-convicted, also with getting a work placement and getting re-convicted. I also found a weak (non-significant) relationship between age of individuals and getting re-convicted. However when I look at the summary data for the saturated model, the significance of my co-efficients seems to suggest the opposite to my analysis.

Could someone please explain what I am doing wrong? Or is this okay? I figured this model would be a good place to start and it would support my findings. Is this the correct approach for finding a suitable GLM?

Here is an image of the summary data as well as the factors

Thanks!

statistics

share|cite|improve this question

asked Mar 13 at 18:54

user553480user553480

62

$endgroup$

• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  Just in case you get no response here, try the statistics site after a while. Don't forget to mention the cross-post here if/when you do.
  $endgroup$
  – Lee David Chung Lin
  Mar 13 at 22:05
  

  
  
  
  

add a comment | 

0

$begingroup$

I am working with a dataset from a job center that specialises in finding work placements for individuals with a criminal record.

There are 2 continuous factors (age, number of pre-convictions) and 4 discrete factors (placement, re-convicted, age group (grouped into 3 levels) and number of pre-convictions (grouped into 3 levels).

My exploratory analysis (consisting of Pearson's Chi-squared test, confidence regions, box plots and barcharts) suggested a strong relationship between number of pre-convictions and getting re-convicted, also with getting a work placement and getting re-convicted. I also found a weak (non-significant) relationship between age of individuals and getting re-convicted. However when I look at the summary data for the saturated model, the significance of my co-efficients seems to suggest the opposite to my analysis.

Could someone please explain what I am doing wrong? Or is this okay? I figured this model would be a good place to start and it would support my findings. Is this the correct approach for finding a suitable GLM?

Here is an image of the summary data as well as the factors

Thanks!

statistics

share|cite|improve this question

asked Mar 13 at 18:54

user553480user553480

62

$endgroup$

• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  Just in case you get no response here, try the statistics site after a while. Don't forget to mention the cross-post here if/when you do.
  $endgroup$
  – Lee David Chung Lin
  Mar 13 at 22:05
  

  
  
  
  

add a comment | 

$begingroup$

I am working with a dataset from a job center that specialises in finding work placements for individuals with a criminal record.

There are 2 continuous factors (age, number of pre-convictions) and 4 discrete factors (placement, re-convicted, age group (grouped into 3 levels) and number of pre-convictions (grouped into 3 levels).

My exploratory analysis (consisting of Pearson's Chi-squared test, confidence regions, box plots and barcharts) suggested a strong relationship between number of pre-convictions and getting re-convicted, also with getting a work placement and getting re-convicted. I also found a weak (non-significant) relationship between age of individuals and getting re-convicted. However when I look at the summary data for the saturated model, the significance of my co-efficients seems to suggest the opposite to my analysis.

Could someone please explain what I am doing wrong? Or is this okay? I figured this model would be a good place to start and it would support my findings. Is this the correct approach for finding a suitable GLM?

Here is an image of the summary data as well as the factors

Thanks!

statistics

share|cite|improve this question

asked Mar 13 at 18:54

user553480user553480

62

$endgroup$

share|cite|improve this question

asked Mar 13 at 18:54

user553480user553480

62

share|cite|improve this question

asked Mar 13 at 18:54

user553480user553480

62

asked Mar 13 at 18:54

user553480user553480

62

asked Mar 13 at 18:54

user553480user553480

62