Search results

  1. hlsmith

    Missing values in Data

    What is your total sample size?
  2. hlsmith

    Adjusting for baseline characteristics

    Correct. MULTIPLE regression allows for multiple predictors. When using it, you can put in the model potential confounders to control (AKA adjusted) for their effect on the outcome. welcome to the forum:) is your name the flip of Luke Skywalker?
  3. hlsmith

    LASSO

    It is a process for feature selection, so model building. Ideally, you use it for figuring out terms using a training set and then fit that model to a holdout set. Do you have enough data to do that? You dont have to do it like that, but it drastically helps for generalizability. If I get time...
  4. hlsmith

    LASSO

    I believe you can only do it with a continuous outcome in SAS if you don't have the DataMiner platform. Is that what you want to do? I haven't used SAS, but I am sure I can help. I have done it with binary outcome a bunch in R.
  5. hlsmith

    Histogram

    Not normal, has assymetry. Seems suspect to some type of selection bias.
  6. hlsmith

    Comments on residuals

    I just meant shave off those extreme residuals.
  7. hlsmith

    Comments on residuals

    Well, run the model with and without the top and bottom 2.5% values (if you can systematically tease them out and know why they are there. Then see how sensitive the model results are to trimming them. You are typically gonna get some weirdness in the tails of the Q-Q plot. Even if you simulate...
  8. hlsmith

    RESOLVED but still open for discussion: When to transform and when to use non-parametric tests?

    @lucyd123 - there is no attachment. I may differ a little from @Miner this time. I would normally go for transformations first. Transformations can result in more interpretable results for people, given you interpret them correctly. Skewed data are common and not necessarily the results of...
  9. hlsmith

    Logistic regression vs. cox proportional hazards model-which model to choose?

    Cox would address unequal follow-up times and loss-to-follow issues. In addition, you would be able to examine the outcome across the follow-up time. If all rehospitalizations occurred within 14 days and you just looked at rehospitalization any time within 2-yrs - yes/no; you would not know...
  10. hlsmith

    SPSS 24 error "document is already in use"

    I like @GretaGarbo recommendation to create a local version of file or perhaps shut everything down, including the computer and reopen everything and try again!
  11. hlsmith

    Commenting on Effect Size in a paper

    Were treatments randomized? I thought they had a cohen's d for all measures (conversion)? Why cant you just use the differences with 95% CI's? Equal sample sizes?
  12. hlsmith

    Test question

    I always go with 1 when starting a project. AKA a literature reveiw, has the question already been definitively answered, and can we improve upon the current state of knowledge, if not, dont proceed.
  13. hlsmith

    Cohort study design question.

    When and how is lead exposure measured?
  14. hlsmith

    CONTINGENCY TABLE

    Also, can you post the contingency table?
  15. hlsmith

    Which test to use? Comparing population with subset from this population.

    Many people feel compelled to conduct formal tests for trivial/descriptive statistics. In this scenario, I am not sure a test is needed, since any difference is just based on the prevalence of males in the full population and the two groups are not independent. For all intents and purposes, just...
  16. hlsmith

    Which test to use? Comparing population with subset from this population.

    Different from women or the whole? The whole includes them, so depending on how many men there are the whole is just a weighted version of the subsample. If you are comparing the genders, a chi-sq or Fisher's exact test can be used, the latter for small samples. If this is just a binary...
  17. hlsmith

    Log transformation and multiple imputation

    I would look at box-cox transformations, which usually just have you add a constant to the variable. If you are using OLS for imputation, I would do it with and without transformation and see what the residuals look like in those models. The residuals are going to likely dictate if a...
  18. hlsmith

    Logistic regression curve looks like linear regression

    And the "plot" thickens to a plasma like state!
  19. hlsmith

    Log transformation and multiple imputation

    Can you describe this in more detail, "OLS/predictive mean matching 10 times"? This can go 3 ways, first off - why does the variable need to be transformed, is it highly skewed? And there is a chance that the skewness may impact the final model but not the imputation model, vice versa, or...
  20. hlsmith

    Log transformation and multiple imputation

    Not clear, so you are using ordinary least squares (OLS) for the imputation, yes/no? Also, to confirm, you say you are using multiply [sic] imputation. Is that correct, too? So you are using OLS 'm' times?