Projects: --- ADA papers (6) My writing project would be my ADA report. I have a whole draft of my ADA report ready but I am not entirely happy with it right now and I think it can be improved. My project is my second Heinz paper, in which I will review statistical arguments made in the medical and legal literatures about the diagnosis of Abusive Head Trauma or AHT (formerly called Shaken Baby Syndrome). At the moment I want finish my ADA report on predicting the flu but may want to adapt that to a paper I plan to use my ADA report as the writing project for this course. I have a full draft which I hope to refine, and I also am planning to try and publish versions of it in both a statistics journal and a psychology journal, which I also hope to work on throughout this class. I plan on doing further work on my ADA report. My writing project would be my ADA report. I have finished the analysis and have a whole draft at hand. But I feel that is only remote from a publishable paper (assuming the material is enough for a publication). ADA Report --- Thesis Chapters (1) I'm working on a chapter for my thesis that involves comparing various methods for choosing models for social network data. Parts of this chapter may also be used for a paper, though my primary goal is to use it as a thesis chapter. --- Thesis proposals (1) Thesis proposal. --- Journal Articles (4) I'd like to complete a manuscript of a methodological journal paper based on my research work. I have bits and pieces written out which I want to complete as a publishable paper. A journal paper I am writing a paper for a journal. It is about comparing two decoders of brain activity in motor cortex. I have an early draft and most of the analysis is complete. It would be very nice if I can improve my draft over the time of this course. I am going to write a journal article detailing the synthetic populations we have been generating for the MIDAS group. We have some preliminary drafts written down, but we want to polish it off and be able to submit it soon. ----------------------------------------------------------------- Purposes & Audiences: --- My Advisors At this point the audience is my ADA advisers. The purpose of the current draft is to summarize my analysis of several aspects of the problem and present tentative results. If we find more promising results, it would be great to turn the report into a paper. --- A Generic Audience The purpose of my project is to convey technical statistical details and interesting findings of my model to a statistically inclined audience. I want to successfully communicate what I have learned and implemented over the course of my ADA. I'm writing for other statisticians. I think people in related areas are my readers. Learn to communicate my research achievement to statistical community. I'd say I'm generally writing for a statistics audience, but I would also like to share this work with the psychology community as well. The statistics included are very psychology-oriented, so I would like the psychology theory to be accessible to statisticians and the statistics theory to be accessible to psychologists as they read my paper. I'm trying to improve the statistics currently used to inform criminal justice policy, so it's very important that I'm able to convey why the current methods aren't adequate and why my methods are stronger. We have developed synthetic populations for various countries, and we want to publish a paper detailing how they work, and how to use them, so they can start to gain some traction in the community. The target audience would be primarily researchers using agent based models to study infectious disease. I want to communicate ideas about how populations of neurons might encode arm movements. My audience is the computational neuroscience community. We don't have a final choice for a journal yet. --- Mission to tell the world something important My ADA project is on a viral disease called Chikungunya and the report talks about predicting the number of cases of Chikungunya in the Americas in the future. I also plan to discuss about the problems that I had to overcome due to the quality of the data available to us. This is important for epidemiologists and statisticians who work on similar diseases. The predicted number of cases in the future could also help people in health care prepare for Chikungunya better with different treatment methods. The purpose of this paper is to critique previous work and provide clear and correct statistical guidelines for i) physicians who need to diagnose whether a specific injury was due to child abuse or an accident, ii) attorneys who are prosecuting or defending an adult who was accused of inflicting child abuse in the form of AHT, and iii) statisticians or other researchers who seek to analyze the data available on the clinical features of AHT and the likelihood of abuse. The report would be directed to astronomers mainly. We applied some not too sophisticated methods to astronomy data to get more desirable results than astronomers previously did. It extracts much more information and gives more powerful results compared to the method previously used by astronomers. We think the method, if applied in more similar analyses, would be helpful. My thesis is focusing on cross-validation methods for network models, and the chapter has two primary goals. The first goal is to convince the reader that cross-validation out-performs alternative methods, such as BIC and modularity maximization, among other model selection criteria. The second goal is to show what cross-validation procedures perform best for specific generative models. For instance we choose the number of folds and the method for assigning folds whenever we perform a cross-validation routine. ----------------------------------------------------------------- Personal: --- The good I like writing my ideas as I think through them. I like writing because it lets me share what I've done/learned. I like the idea of communicating my research. I like when a difficult concept can be communicated clearly and simply, often aided by key graphics. The aspect I like the most while writing a report in Statistics is the usage of tables, plots and other visualizations of the data or the results. I feel like these are the easiest to explain and interest the audience. I like that technical writing does not have to be long to be of high quality. It can be concise and as long as the ideas are good, it can be of good quality. I enjoy writing accurately and technically. I enjoy technical writing because it's a way to share my work with a much broader audience than a presentation or a poster session. It's much more of a standalone product and it's easier to give a full picture of everything that I've done. I enjoy the challenge of being able to communicate ideas from the statistical world back into reality, as well as seeing if I can explain a technical concept to other statisticians/scientists. It also provides a structured way to fill in details in my thinking (as in I left a placeholder paragraph for this topic so I should think about it some more). What I like about writing (though it tortures me sometimes) is that it made me clear up my thoughts and state the analysis process from bottom up. It somewhat helped me find the real significance of my work. --- The bad My main struggle is coming to a complete understanding of the technical concept I'm trying to communicate. I find explaining methods used to obtain the results the most difficult task. I find it very challenging to explain statistical models in an interesting way so that readers would stay interested and understand what I am trying to say. The most challenging aspect for me is making sure that things that seem *obvious* to me aren't inferential leaps to someone who isn't as familiar with the topic. I tend to have a difficult time imagining what people will have trouble understanding. I struggle most with finding the line between explaining things that are obvious and leaving out things that are essential for the readers understanding. The challenging part is I am not too familiar with my readers, sometimes it wasn't obvious what to assume about my reader, especially when I'm writing for someone outside statistics. I find transitioning from a technical rough draft to a more formal (shareable) paper most challenging. I find it challenging to write things up! Maybe this is a silly way to say it, but I have had several ideas and results for papers, and I haven't been able to write them down as papers for some reason. I am hoping that this course makes that process easier so I can actually show some of the work I've done. What I don't like about writing is how much time it takes. I have always found writing in general to quite challenging. I have found that technical writing as gotten a bit easier with experience. --- How I want to improve I am hoping both to improve the quality of my statistical and technical writing as well as making it less of a chore. I expect that as I progress as a statistician, regardless of the direction my career takes writing will become an increasingly large and import part of my job. So it is a skill I would like to focus on developing further. More specifically, I struggle clearly explaining the steps I took to create a working project and want to clearly explain these steps so a reader can replicate it. I would like to learn more about how to write logically and clearly. I feel that clarity and concise are most challenging for me. A lot of times it's challenging to put down in words why I've made certain decisions, or why a specific result is important. I find the most challenging aspect is being concise while still conveying everything I want. What happens to me is that I've learnt English by taking class at school and nothing I've done in the past was for statistical purposes. I have a basic knowledge on writing in general and even less knowledge on writing technical papers. For instance, I have problems to distinguish formal and not formal language. When I write for my advisor for preparing my meeting, I tend to write key things mainly aimed to discuss results together, but I never write the whole story because I know neither how to organize the small report, nor what language I should be using. The most challenging aspect for me is being patient enough to work through constant revisions and criticisms of my work. I would like my paper not just to be able to convey concepts to statisticians, but also to a diverse audience that is not familiar with statistics. I want to be able to write in a way that is systematic and easy to understand, but also efficient. --- Writing Habits I maintain a technical report for most of my work, but write more formally only when I have substantive part of the problem I'm working on figured out. I write every other day, primarily just summarizing what I'm reading, so it's usually not very polished. I like writing pretty much once a week. But for my project and statistical purposes in general, I now write when I have to. I was writing a closed blog, for updating my thesis results, but I stopped because it was taking a lot of time and I was continuously struggling with language issues and stopping my research. I usually write about once a week and find it very difficult to sit for a long stretch of time. So instead I plan to work on writing three days a week this semester and write for short durations of about an hour or so. When I'm working on a project and have enough results to begin writing, I'd say I write once or twice a week. I like to have at least 4-6 hours to sit down and work on it. I find when I try to write every day, I don't leave myself enough time to do all the editing that I want and sometimes forget to fix it later. Currently I'm writing at least once a week for this project, though I would like to be writing more frequently. I probably write up my work once or twice a month, which is an area where I want to be more diligent. Once a week. I definitely still fall into the category of writing only when I "have" to. I don't write often. Most of the time I prefer to work on an outline of figures and delay writing until the very end. I typically write in batch-mode; I'll keep terse notes while I'm working on something and when I "have" to have something done or I have a lull in actual work I'll write up sections. During my ADA projects I am very much a binge writer. I am doing the "writing schedule" attempt as described in the book How To Write A Lot, trying to sit down and work on writing at fixed time every weekday. I hope to become a daily writer from now on. -----------------------------------------------------------------