[Japanese] | Education | [Top]

Health Sciences Common Special Lecture IV, VIII

[http://minato.sip21c.org/ebhc/index-e.html] Latest update: 6 August 2019 (Tue)

Archives: 2018, 2017, 2016, 2015, 2014, 2013, 2012

1. Aims of the lecture

For properly conducting health related research, basic knowledge of research design and statistics will be given.

In principle, the lecture is done in Japanese. Foreign students may ask the lecturer individually if necessary.

The text written in Japaneses [pdf format], latest version is on 24th July 2019 is freely available (Comments are welcome).

Caution! There is no lecture on June 12 and July 10. Lecture for experimental data anaylsis is given at Saturday afternoon on July 27.

2. Program

1. The very basics of health research (Apr 10, handout in Japanese)
2. Basic terms in health/epidemiologic research (Apr 17, presentation, exercise: If you are absent on this day, you'll have to submit the answered exercise via e-mail.)
3. Sample size issues (Apr 24, handout in English) (cf. E-learning course on ICRweb is useful.)
4. Research design (mostly experimental study design) (May 8, handout) [A review paper pdf for crossover design]
5. Data entry, missing data, graphical check (May 15, Excel format sample data)
6. Hypothesis testing, SD and SE. Test of the location difference (May 22, handout in English)
7. Test of proportion difference, Cross tabulation and its analysis (May 29, handout)
8. ANOVA and Multiple comparison (June 5, handout)
9. Correlation, Regression and its application: Analysis of Covariance and Logistic Regression (June 19, handout)
10. Geographic information data analysis and spatial epidemiology (June 26, Special Lecture given by Dr. Susumu Tanimura)
11. Repeated measurement: Comparison of location parameters among 3 or more groups with corresponding data, Kappa statistics, McNemar's test (July 3, handout, Bhapkar's test code)
12. Evaluation of performances of diagnostic testing: ROC and Bland-Altman plot and the method to summarize systematic review: meta-analysis (July 17, handout for ROC and Bland-Altman plot�Chandout for meta-analysis)
    Example of meta-analysis: Is sperm concentration decreasing over the world?
    • The first paper written by Carlsen et al. (1992) in BMJ, which includes the data as table. According to the name of the research leader, so-called "Skakkebaek paper".
    • Reanalysis of the data in "Skakkebaek paper" by Swan et al. (1997)
    • Re-reanalysis of the data in "Skakkebaek papar" with newly added data by Swan et al. (2000)
    • New paper criticizing the "Skakkebaek paper" and followers about sampling bias written by Deonandan and Jaleel (2012)
    • New review paper (2016)�CNew cross-sectional study in Denmark (2018) by Prof. Skakkebaek's group.
    • New meta-regression paper by Dr. Swan's group (2017) (In addition, the supplementary material explaining the actual process of data collection and analysis is also available. However, downloading the data requires the password, available from the corresponding author.) (Interview to the first author is available) (Comments from professionals)
    • My web-log critics on "Skakkebaek paper" in 1998 (in Japanese), which has been later published in "Kagaku" journal by Iwanami-Shoten
    • Tab-delimited text file of the data in "Skakkebaek paper"
    • The R code to plot the data above as same graph as "Skakkebaek paper"
13. Survival analysis (handout in English�Cref. in Japanese) (July 24)
14. Experimental data analysis (July 27 [Sat], Given by Dr. Tokuhara, 13:10-16:20)
15. Mini test (July 31, If you cannot take this exam, please inform me in advance.) (exam with answer in pdf (in Japanese))

3. Evaluation

Based on the mini-reports and mini-test.

References

• Machin D, Campbell MJ, Walters SJ (2007) Medical Statistics, 4th ed., Wiley
• Lang TA, Secic M (2006) How to report statistics in medicine: Annotated guidelines for authors, editors, and reviewers. 2nd ed., American College of Physicians.
• Peacock JL, Peacock PJ (2011) Oxford handbook of medical statistics. Oxford Univ. Press
• Power and Sample Size (Vanderbilt University)
• EZR (Jichi Medical School)
• The lecture note for medical statistics at Gunma University written by Minato Nakazawa
• dfamefs.R: R code to calculate d-family effect size
• pooledr.R: R code to calculate pooled correlation coefficient using Fisher's Z transformation and weighed average

Notice to cite and link