I guess I can contribute some information about the 3000 modules. Everything would be based on when I took them, so it may not be the same now.
ST3239 Survey Methodology seems to be offered once every 2 years as the last time I saw it was 2 years ago. The materials are quite easy to understand and most of the problems involve calculations. I guess you could consider it to be one of the easier stats module around, making it a fine choice to take.
ST3232 Design and Analysis of Experiments involves ANOVA and experimental designs. Although not required, having prior knowledge of ST3131 is an advantage. R is also used.
ST3241 Categorical Data Analysis includes contingency tables, odds ratios, likelihood ratios, generalised linear models and logistic regression. SAS is used. Overall I found it to be doable and not too difficult.
I didn't take ST3242 Survival Analysis but from what I heard it is quite OK.
ST3244 Demograhic Methods is an interesting one. There is not a lot of statistics involved. It talks about growth/mortality/birth/death/fertility rates, life tables, survival probability and population projections. If you have taken Acuturial Statistics last sem, the chapter on survival models would be familiar. I recommend it as I enjoyed it and found it to be applicable to real-world situations. The lecturer(same for this sem) is good too.
Like I said in my last post ST3236 Stochastic Processes 1 is offered by the maths department this coming sem so I think it would be a better idea to take it next AY sem 1 when it should be offered by the stats department. It is really tough and some abstract concepts are dealt with. But then again who knows, the maths department might teach it better.
Take a look at the past year materials for those modules at the usual source to see more of what they are about.