Search results

  1. staassis

    Simple Question

    Anything can happen due to randomness. This is the case of binomial distribution. Please check out the link. Having that said, because the sample size is big (3,000) the sample proportion of customers making $10,000 will be close to 1/30. The Law of Large Numbers will work.
  2. staassis

    Mediation conditions

    Yes you can. The issues of statistical significance and effect size are separate. Even though certain bacteria are tiny, proving their existence is important. They matter.
  3. staassis

    Panel regression: Deciding on fixed effects and random effects - Hausman

    Hausman test is only one of criteria when comparing the fixed effects framework to a random effects framework. You should also check how one approach fares against the other in terms Akaike Information Criterion, Bayesian Information Criterion, goodness of fit or something similar. If fixed...
  4. staassis

    Which ANOVA to use in SPSS?

    In general, this is a set-up for Multivariate Analysis of Covariance (MANCOVA). The dependent variables are grammar and oral proficiency, the factor is education and the covariate is years of residence. However, how you go about the actual analysis depends on 1) whether the residuals are...
  5. staassis

    joint probability distributions

    Since the joint distribution is symmetric in X and Y the marginal distributions of X and Y are the same. You are given X's marginal density f_x(x). So mu_y = mu_x = Int x * f_x(x) dx
  6. staassis

    Correlation analysis: spearman or pearson

    The displayed relationships are fairly linear. If fact any non-linear terms, when added to the model, may not come out as statistically significant. If you try both, Pearson and Spearman, they are likely to deliver the same significance status. In a situation like yours they usually agree. This...
  7. staassis

    HELP

    Whether the intervention is effective or not is determined by p-value of the respective Wald test, or likelihood ratio test. Only if the coefficient is statistically significant you look at the estimated eta square. It is an extension of the idea of partial R square. Eta square quantifies the...
  8. staassis

    What test should I do?

    Is this everything the professor gave you? To decide on the test you need to explore the distributional properties of the data. And for that you need more information. For example, having a value for each participant would suffice. On the other hand, seeing only ranges and medians would not be...
  9. staassis

    Which Statistical Tool do I use?

    Which optimization problem? Define "efficient". Again: what you are trying to achieve? What is the objective function (which you are optimizing)?
  10. staassis

    Which Statistical Tool do I use?

    The answer depends on what you mean by "closest" and what you are trying to achieve. Please elaborate.
  11. staassis

    Correct for multiple comparisons with two time points and contingent relations

    The number of sub-analyses is the number of p-values you generate. Does not matter where. If 1] for each of the correlations, you calculate a p-value to decide whether the correlation is significant and 2] all these correlations are used to shed light on the same research question sum all...
  12. staassis

    Correcting sample response data (Likert scale data)

    Never throw away the data. Just recognize that the distribution is far from Gaussian and use appropriate statistical methods, nonparametric and parametric.
  13. staassis

    Correct for multiple comparisons with two time points and contingent relations

    The number of comparisons is the number of p-values you generate for inference. It is never a good idea to generate too many p-values.
  14. staassis

    Comparing multiple dependant correlation coefficients

    I did not understand your message fully but this seems to be a set-up for Repeated Measures Multivariate Analysis of Covariance (Repeated Measures MANCOVA). If the assumptions of plain vanilla Repeated Measures MANCOVA are not satisfied, you can run MANCOVA based on bootstrap.
  15. staassis

    Statistical Tests

    You are using a time series model to estimate Wave Height. Therefore, you need to run time series diagnostics as part of the overall goodness-of-fit diagnostics. 1] Diagnostics based on residuals: define Residual = Wave Height - Predicted Wave Height. 1.1] Estimate autocorrelation function...
  16. staassis

    Interpret Linear Discriminant Analysis and visualization

    LDA requires that all the features be scale variables, not categorical or ordinal. Moreover, it requires that all of the features had joint normal distribution.... I know. Moreover, LDA requires that the covariance matrix of features in class A be the same as that in class B. That is what I...
  17. staassis

    Interpret Linear Discriminant Analysis and visualization

    You do not have to scale. However, the variability of the features and their correlations must be the same in each class. If that assumption fails, LDA may misfire. There are other situations in which LDA misfires and is suboptimal to other classification methods... I'm afraid there are...
  18. staassis

    urgent help required - data analysis of time course

    The p-value from the omnibus ANOVA test may be larger or smaller than the p-value for a t-test comparing just two particular groups. What matters is that there is only one ANOVA p-value but many t-tests (multiple testing). And none of those t-tests summarize the whole dynamics in a statistically...
  19. staassis

    Interpretation of second difference of a predictor

    To some extent, the second difference is measuring whether the companies have generated more casfflows than last year or less.