Sumários
Lesson n.º 12
7 Outubro 2019, 10:30 • Maria Helena Mouriño Silva Nunes
Duration: 2h30m
Lesson n.º 11
19 Setembro 2019, 10:00 • Maria Helena Mouriño Silva Nunes
Duration: 2 hours
Seminar:
Health Technology Assessment: the role of MEDTRONIC.
Lesson n.º 10
18 Setembro 2019, 14:00 • Maria Helena Mouriño Silva Nunes
Paired Comparisons: Conducting a hypothesis testing for the difference between two population means based on paired samples (t-test and large-sample z-test). A summary of the main steps.Application to the database "Cholesterol levels after heart attack": comparing the cholesterol levels, in terms of means, 2 and 4 days after the heart attack. Detailed analysis of the outputs obtained from the SPSS.
Confidence Intervals. An overview. Application of these concepts in the database "Cholesterol levels after heart attack".
Lesson n.º 9
17 Setembro 2019, 14:00 • Maria Helena Mouriño Silva Nunes
A brief revision of the most important results studied in the last class: comparing individuals who fasted before the blood tests to whose who did not fast, in terms of the glycemia levels (mean values). Graphical comparison by means of the parallel box-plot as well as running the independent t-test. Transforming the p-values obtained from the outputs of the SPSS, based on the t-tests, when the alternative hypothesis under consideration is not bilateral but rather a left (or right) unilateral test.
One-Way Analysis of Variance to compare the means of three or more populations. Establishing the null and alternative hypothesis. Assumptions of this methodology. ANOVA Table and its components. Test statistic and respective p-value. A brief reference to Multiple Comparisons when the null hypothesis of equality of the means is rejected.
Levene's test to evaluate the homogeneity of the underlying populations. How to proceed when this assumption is not satisfied. A robust version of the analysis of variance in the case of heterogeneity: Welch statistic; Brown and Forsythe statistic. An example: comparing the mean levels of glycemia based on global cardiovascular risk (4 groups).
The analyses were carried out using the database "Evaluating risk for cardiovascular diseases in the Portuguese population". In terms of software, we used the spreadsheet Microsoft Excel and the IBM SPSS Software.
Lesson n.º 8
16 Setembro 2019, 14:00 • Maria Helena Mouriño Silva Nunes
Summarising the main points of Lesson 6, with special emphasis on the different steps to conduct a hypothesis testing. Explaining in detail the concept of the p-value, with some examples.
Hypothesis testing to analyse a single mean: t-test based on the Normal distribution, and respective approximation for large samples. Analysing step-by-step the whole procedure. An example based on cholesterol levels (data obtained from the database "Evaluating the risk for cardiovascular diseases in the Portuguese population").
Using graphical representations to make an empirical comparison between two groups of independent patients. Histograms and parallel box-plots.
Hypothesis testing to compare the parameters of two independent populations. The main steps: - The F test to compare the variances of the two populations (populations homogeneous versus heterogeneous); - The independent t-test to compare the means of the two populations.
Application of these methodologies to the database "Evaluating the risk for cardiovascular diseases in the Portuguese population". Comparing the cholesterol levels of individuals with and without diagnosed diabetes (the group with disease versus the control group).