Can somebody give me a good way to make an interpretation Demographic factors [age,gender,workplace,years worked in a ED,work regime] (independent) and risk perception to violence and worriedness about violence(dependent).
Dear, need to apply logistic regression making dependent variable as binary and look for the ODDs raio(expo(b)in SPSS) . In order to find out asociation individually ie crude odds ratio for each do it one by one... the highest odds ratio will tell u the the strength of association btwen the variables.also look for P value simultaneously.if its value is
in addition to what is written, I suggest you pay attention to the value of the confidence interval (select option "95% confidence interval"). the lower and upper 95% CI will provide information about the range within which it can vary the relative risk estimate (odds ratio through - exp (B)).
the confidence interval gives you information on the statistical significance of association (such as p value) when cmprende the value 1.
Also, note that there will be gender differences in risk perception to violence and worriedness about violence(dependent). If there is a significant difference in these two depend. v. it is better to do stratified analysis (i.e., run separate logistic regressions for male and female. Usually predictors are different for males and females as perceptions vary greatly between genders. Hope this helps.
Pearson's chi-square value is different from the likelihood ratio because Pearson is more appropriate where variables are continuous with a distribution and results are evaluated by reference to the chi-squared distribution. It tests a null hypothesis stating that the frequency distribution of certain events observed in a sample is consistent with a particular theoretical distribution.
Pearson's chi-squared is used to assess two types of comparison: tests of goodness of fit and tests of independence.A test of goodness of fit establishes whether or not an observed frequency distribution differs from a theoretical distribution.
A test of independence assesses whether paired observations on two variables, expressed in a contingency table, are independent of each other—which I think is your case, right?
where as..
a likelihood ratio test compares the fit of two models, one of which (the null model) is a special case of the other (the alternative model). The test is based on the likelihood ratio, which expresses how many times more likely the data are under one model than the other. This likelihood ratio can then be used to compute a p-value, or compared to a critical value to decide whether to reject the null model in favor of the alternative model. When the logarithm of the likelihood ratio is used, the statistic is known as a log-likelihood ratio statistic, and the probability distribution of this test statistic, assuming that the null model is true,
so Pearson chi-sq tests the distribution and the likelihood ratio tests the fit of the model.
to answer your second question where you said using the same variables in crosstabs and in logistic reg your values are different ..
I ran some analyses to see what you are referring to.. I ran crosstabs and also logistic reg. with the same variables
There is no difference in the values I got and they should not be different. I got the same values for both Pearson's chi-square value and the likelihood ratio in crosstabs and logistic.
I think you have a coding issue. Make sure that when you are doing crosstabs use 1 for yes and 2 for no for both outcome and the predictor variable. Use outcome in the column.
Then when you are running logistic, recode the outcome variable keep 1 as yes but recode 2 no as 0
and keep the predictor as is. Then you will get the same values.
Yes, it is important that you keep your coding consistent, particularly if you are using SPSS. SPSS has internal auto coding that changes our specified coding. I am glad you got that out of your way.
Now, concerning your OR being big, what do you mean? Do you not expect to be that big between the groups you are comparing? Is your design cross-sectional or case-control matched or unmatched? If it is matched you need to conduct different analysis. Regardless, the interpretation of 3.55 will be that compared to your reference group whatever the other group is, te relative odds are 3 1/2 times higher which is significant (p
In addition to all valuable comments on logistic regression analysis, you need to pay attention to the values in column B which represent the extent to which the value of that independent variable contributes to the value of the dependent variable. You can then put the first, second , third predictor according to B with being attention of CI and P value
Hello all, your comments here have also been useful to me. However I have a question to ask about interpretation. At Chi-square, the variables that were significant were inputted into d logistic regression model in order to determine the co variates that predicted my dependent variable. the logistic regression however showed that the level of significance was greater than 0.05 which suggests that non of the covariates remained significant predictors of my dependent variable. My question is that what is the best way to interprete this result? should I completely discard the odds ratio since my p was not significant? Thank you all