Recent content by hlsmith

  1. hlsmith

    How much of Science is based on Statistics

    Since I got a like from @ondansetron - I thought I would also mention that I read last night Yule was able to grasp and explain autocorrelation in time series using physics (e.g., pendulum swinging). Also, since you all were referencing extreme p-values and collider research, I reviewed an...
  2. hlsmith

    Hi everyone!

    Purge those question into posts and we will hope that someone can answer them. Welcome aboard!
  3. hlsmith

    Hazardratio statement, interaction in Proc Phreg (cox-regression)

    Alright, I will take a look on the web... This is what I found, seems achievable using the contrast statement in this document. Post the solution you come up with or if you need more help. https://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/viewer.htm#statug_phreg_sect034.htm
  4. hlsmith

    Hazardratio statement, interaction in Proc Phreg (cox-regression)

    Yeah, I would imagine the HazardRatio statement is needed to specify the desired estimate. Perhaps look towards proc logistic examples to see how they solved a similar question, I am guessing with an estimate or contrast statement. Sorry I don't have time to be more helpful. :(
  5. hlsmith

    Hello everyone!

    Welcome aboard, i have two daughters as well! We look forward to your questions and contributions.
  6. hlsmith

    calculating the effect size

    Predicted probabilities and accuracy are two model metrics used regularly.
  7. hlsmith

    Can I use Multiple Regression on my data?

    Seems possibly reasonable to do. The important thing is that you meet the model assumptions.
  8. hlsmith

    calculating the effect size

    @noetsi I agree with you, but I believe some people do deem these partial R-squares as effect "sizes".
  9. hlsmith

    Bootstrapping in SAS

    @noetsi - lets see the reference to using bagging with time series. I start what seems like it will be an intro to time series in 10 days. So I will learn a little more and hoping catch-up to your in your time series lingo. Cheers!
  10. hlsmith

    Revman 5-Generic inverse variance method

    Not overly familiar with Revman, but yes the 95% CI values are also invariable, meaning you can also use their inverse. Whenever I have a question like this, to confirm the validity, I will just run a model both ways using an available dataset to show that it works - for reassurance purposes...
  11. hlsmith

    calculating the effect size

    Not familiar with SPSS, but this should get you in the right direction: https://www.theanalysisfactor.com/partial-eta-squared/
  12. hlsmith

    Spearman Correlation or Kendall

    Are you just caring about the first and last columns?
  13. hlsmith

    Spearman Correlation or Kendall

    Can you provide an actually faux example of your data? Also, what is your sample size? Thank you!
  14. hlsmith

    calculating the effect size

    @Eleni A, In the social sciences the tend to use partial R-square as an effect size. So if I have a model: y = X1 + X2, I could generate the amount of R-square explained by X1 and X2. I can also slap confidence intervals on these values. They are called something like omega or epsilon partial...
  15. hlsmith

    Trial sequential analysis of noninferiority randomized control trials

    Hmm, interesting topic. I have not done really any of these analyses before. For non-inferiority, I usually double my alpha, since it is actually one-sided - this is when I use traditional null hypothesis. But it seems you are looking to redefine the null value (constant you are using for a...