Poisson Regression Analysis in SPSS with Assumption Testing

[@miranozmi3978]

thank you Dr. Todd. Took me 3 times watching this before I understand what you explained.

[@kathystansbury771]

Hi Todd - Thank you for the really great video and step by step analysis of the assumptions tests. I've used it several times. One comment that I have is that it would be great if you showed an example in which the data did not meet the dispersion assumption. Cheers, Kathy

[@nikolaostsigilis1580]

Excellent video, to the point. Contains all the necessary information. Well done.

[@DrGrande]

You’re welcome!

[@florahuangsg]

Thank you for your excellent videos.

[@abat5991]

Thank you for the excellent to the point presentation

[@eliza6438]

And if er have in the Smivoc test .05 statistically significant and not follow poisson distribution. What choice should we set in the glm?

[@gunalbilek9675]

Great work Todd. Cheers

[@priyathgregory9055]

I have a situation where the omnibus test is non-significant. How can I work around that?

[@taritakuma8895]

How can I get an incidence rate or incidence rate ratio with SPSS for Poisson regression?

[@facoffee7036]

What if I have only one series of data, e.g. the frequency of occurrences of binned values? In this case I think I only have the predictor variable, correct? I need to estimate the Poisson's lambda and can't figure out how to do it

[@Dennis-J316]

Thanks very much for the video, very useful, I have a question: what if the distribution of counts in a depended variable not follow Poisson distribution? Like in KS Test the p value is less than .05 and the Mean and Variance are too differ from each other. what need to do in this case?

[@annashivute5978]

Hi there Dennis :) came across the video today. I have the same issue. Did you perhaps find the solution to it, and if you did, could you kindly assist and share it with me. Thanx :)

[@quixoticanalytics9717]

@@annashivute5978 I believe a negative binomial regression is then used and you can then adjust it and check the AIC

[@annashivute5978]

@@quixoticanalytics9717 Thank you very much

[@florahuangsg]

@@quixoticanalytics9717 Thanks so much.

[@susanv.dbroek9771]

A very helpful video but I still have a question. I want to know what the effect of self-esteem (scale variable) is on the number of celebrity influencers someone is following on social media (count variable). I understand how I should run this test in SPSS. However, I do not know how to include my control variables. I want to control for: the age of the respondent (scale), the education of the respondent (ordinal), the gender of the respondent (nominal) and nationality of the respondent (nominal).

[@GoogleUser-xs4su]

Thanks for the video. Could you please you or someone else here answer to the following question? When we interpret the results (the parameter estimates box), should we say that adjusting for the covariate (i.e. functioning), male attended 14.1% more meetings? I'm not sure how to keep the covariate in the picture. Thanks so much!

[@TRKovacs]

Thank you for your very helpful videos. I was a bit confused by your explanation of the results for your nominal variable (gender). I'm using count data to study developmental skills and tried to follow your example for coding two age groups as a categorical variable. If I code my age groups as 0 = younger, 1 = older, my exponentiated beta value is less than 1 because SPSS calculates change from the older group to the younger group. In order to see developmental change moving from young to old, I need to enter 0 = older, 1 = younger. If there is some logical relationship between categories, you need to be careful how you number them.

[@khaledabdu2]

where is the data file please could you upload that

[@serdecsomalia3669]

THANK YOU INDEED DR. YOUR PRESENTATION IS HIGH QUALITY AND USEFUL BUT THE VOICE IS RECORDED VERY LOW.

[@i-am-oi]

Please am working on a research that deals with certain close ended which has to do with rank and another has several incentives workers may choose as either very important, just important or not important the last one requires each individual to choose between monetary and non-monetary incentive. How do i go about this using spearman with the help of spss. Pls kindly assist

[@johnsungwachanasila3128]

Thank you, Dr. Grande, for your wonderful presentation, which was well detailed to a level which anybody with interest could clearly understand. Now Dr, I have some data which I would like your input indicating the right approach to analysis. It is medical data based on counts taken on several variables.

[@mrodriguezgallardo]

Hello , can you tell me how to download the data?

[@amishadykiswaga1077]

very useful. am asking you have used the scale dependent variable, what if it is dichotomous?

[@kakunika131]

Dear Dr. Todd Grande, thanks for your video is too helpful. But i have one doubt, how can we make when the distribution is too overdisperse? thanks in advance.

[@gunalbilek9675]

Hello Todd. Thanks for the video. Do we use odds ratio of intercept for interoperation? I noticed that multinomial logistic regression does not produce odds ratio for intercept? I am wondering why. Thanks in advance.

[@linliyyee6242]

i have 7864795 cases in total, and when i did the nonparametric test, it appeared to be not available. is there any other method to conduct similar test to prove that they follow poisson distribution beside nonparametric test? Thks for the video.

[@elprofequecorre]

Los vídeos son muy explicativos y claros en sus descripciones, lo único que debería mejorar es el zoom de la pantalla, ya que los textos y números quedan muy pequeños.

[@ugurcansayili9543]

Very helpful! Thanks...

[@etiennensereko1262]

Thank you for this video

[@halloikbeneenpotvis]

So what do you do when value/df >> 1? (Mine is about 24000)

[@aprianideww1559]

Are u have the data?

[@banaanwijers]

Very useful, thanks!

[@MPHEBRiri]

Legend.

[@mehjabeenmusharraf2625]

Great. Data file please

[@stenikoli1]

OBRIGADAAAAAAA!!!

[@lorraineraby2055]

Very helpful! Thanks.