Search results

  1. staassis

    Somebody please help me

    I agree with @hlsmith: use all the data + t-test + quantile regression. Precisely because the group sizes are so big, the Central Limit Theorem guarantees validity of t-test. You may also run two-sample Kolmogorov-Smirnov test to see if the distribution in the test group is the same as that in...
  2. staassis

    Timeseries Multiplicative Triple Forecasting model

    General observation: Excel is a primitive package. There is much more time series functionality in packages like R, Matlab, Stata and EViews. This functionality will help you run diagnostic tests on the estimated model and extend the model accordingly. In particular, you may want to focus on the...
  3. staassis

    Trying to determine attrition

    This seems to be a case of survival analysis in discrete time. In particular, the Kaplan-Meier estimator of the survival curve, alongside the corresponding confidence band, may be of use.
  4. staassis

    Help with data analysis plan for dissertation

    To the best of my knowledge, the PROCESS add-on assumes normally distributed residuals. If they are not normally distributed and the sample size is modest, you are better off estimating various models with interactions using standard SPSS functionality (standard SPSS edition). In many "linear"...
  5. staassis

    Which Statistical Test To use?

    Stata is fine.
  6. staassis

    Which Statistical Test To use?

    For each dependent variable, you can estimate ARMAX time series models. Before entering the dependent variable (e.g. Net Interest Income) and the independent variable (Interest Rate) into a particular model, you have to transform them to make them stationary. In other words, if a given variable...
  7. staassis

    Which Statistical Test To use?

    What are your research questions?
  8. staassis

    Determining sub-period probability from overlapping temporal probabilities

    This probability cannot be determined without making extra modeling assumptions. It is customary to model the number of rare events like thunders in a given time window via non-homogeneous Poisson process.
  9. staassis

    Questionnaire Data - which stats test

    Focus on the beginning of the first post, where various tests are listed. Then you can look them up separately in Wikipedia or an introductory statistics textbook. Is there a Layman's guide? Yes: an introductory statistics textbook. One example: Freedman, D., Pisani, R., & Purves, R. (2007)...
  10. staassis

    Fit review for multiple systems

    In general, you can perform estimation in a loop using R, Matlab or Stata. And then, you can aggregate residuals over systems for each configuration and compare the 6 configurations using a measure like BIC (Bayesian Information Criterion). However, in your situation 8-9 data points are not...
  11. staassis

    HW Question: Log-Likelihood Function

    Plot the log-likelihood that you have derived and see if you notice anything unusual. Visualization is almost always an important step in diagnostics.
  12. staassis

    Repeated measures in SYSTAT 12

    Yes. You can look up my web-site in my profile and write to me there. Regards.
  13. staassis

    Questionnaire Data - which stats test

    Please see this thread on testing relationships between binary variables. It will answer your question partially.
  14. staassis

    Regression with no intercept - null assumption for

    If you are discussing a test for the slope coefficient (b1), then in a linear regression with no intercept H0: y = epsilon, H1: y = b1 * x + epsilon, where epsilon is a random error (residual).
  15. staassis

    One way ANOVA or Two Way ANOVA?

    Yes, that is true. But is the cell material different between the days? Please check with the professor.
  16. staassis

    One way ANOVA or Two Way ANOVA?

    Are different drumsticks tested on different days or are they the same for each preservative?
  17. staassis

    Is ANOVA right for me?

    Your experiment is a case of Repeated Measures ANOVA if the residuals are normally distributed. The normality can be verified using Kolmogorov-Smirnov, Shapiro-Wilk or another test. If the residuals are not normally distributed, you should use a nonparametric version of Repeated Measures ANOVA...
  18. staassis

    Moment generating function

    Use the fact that E[Y^k] is the k-th derivative of the moment generating function evaluated at t = 0.
  19. staassis

    Using Residuals to Make Conclusions

    So analyze the plot for the distributional properties I mentioned (symmetry / asymmetry, tails, etc). To check unimodality, you may use histograms.
  20. staassis

    Confidence intervals for proportions

    No. A CI for 1/r is to be obtained algebraically from the CI for r.