Social Epidemiology - Science topic

This is actually an important question. The answer (at least in my opinion) is an emphatic "yes". Different research designs make different assumptions about the processes being evaluated by using different types of variation in the data. For example, all else equal, a cohort study is usually considered to be stronger than a cross-sectional study because it exploits data variation between cohorts in the same country (city, region or whatever the level of analysis is), whereas a cross sectional study relies on variation in the data between different countries. It is usually the case that there are more "alternative explanations" for differences between countries than between cohorts.

To summarize, do consider different study designs. If many different study designs have the same result, that is often strong evidence of a meaningful relationship. If different study designs have different results, you should think about why that is and what it tells you about the relationship in which you are interested.

For presentation.

Thank you for helping. :) 

Okay, another book that talks specifically about ethnographic coding is The Ethnographic Interview by James Spradley.   This book has clear details and illustrations for coding ethnographic research.

Allison,

Consider . C. Ai and E. C. Norton, "Interaction Terms in Logit and Probit Models."  Economic Letters 80 (1): 123-129.

With good wishes !

I do research on safety and security for human rights activists and I have found the same problem. But, for your case, perhaps you could have a look at the Adult Hope Scale? (http://www.positivepsychology.org/resources/questionnaires-researchers/adult-hope-scale) And more generally there is stuff about the "sense of hope" wich might be relevant for you?

And similarly about risk and uncertainty in everyday life, like:

DOI   10.1080/13698570802380891

Calculate the 95% CIs of the variance (low and high: ie variance +/-  1.96*SE of  the variance) and then plug these values into the MOR formula. The problem with this approach is that you are assuming (with likelihood- based estimation) asymptotic normality to obtain the 95% confidence intervals and that is why many prefer MCMC estimation to obtain 95% credible intervals. One can anticipate that the distribution of the variance will be positively skewed especially if the effect is not large and there are not many higher level units.

See the chapter on the binomial in this

Among the physical determinants do not forget the sound environment of children in the school, at home and in the community, where the child lives

doi:10.1016/S0140-6736(05)66660-3

doi: 10.1093/aje/kwj001

doi:10.1136/oem.59.6.380

doi: 10.1177/0013916503256260

doi:10.1016/j.envres.2015.08.003

doi:10.1016/j.envint.2011.03.017

doi:10.1136/oem.2006.026831

Dear Sander I could find this for you, may be it helps:

Validation of the Chinese translation of the 6-item De Jong Gierveld Loneliness Scale in elderly Chinese

Assessing Chinese Adults' Intellectual Abilities

Diagnostic Performance of Short Portable Mental Status Questionnaire for Screening Dementia Among Patients Attending Cognitive Assessment Clinics in Singapore

http://www.annals.edu.sg/pdf/42volno7jul2013/v42n7p315.pdf.

The prevalence of functional disability in activities of daily living and instrumental activities of daily living among elderly Beijing Chinese

Aamir

You might already have cases of Zika and not know it.  The huge difference between Zika and chikungunya is the percentage of people who get sick.  Actually Zika seems like dengue in that both viruses are unlikely to cause someone to be so sick as to present at a clinic.  If you look at the statistics from some longer studies it seems like with Zika maybe less than 20% of people present.  This and the fact that it seems similar to chikungunya and dengue there was a great probability that Zika was being missed over the last year or so.

Thanks Thomas, I have found one study but can only find a draft publication online. Any ideas how I can track the final version?

My reference for it would be: 

Prota, L & Cucco, I 2014, 'The impact of garment work on commuters' villages in Cambodia', Presented at the workshop on Globalization, Industrialization and Labour Markets in Asia: Essays in Honour of Melanie Beresford, January 2013. Chiang Mai, Thailand.

Suppose

The gold standard formula for converting OR to RR is:

RR = (OR)/(1-Ro+Ro*OR), where Ro = c/c+d, ie, the probability that the non-exposed will develop the outcome/disease.

I assume that in your case the problem is you do not have 'c' and hence Ro can not be calculated.

There is a method to overcome this by some optimisation provided you have the total number of treated (exposed, a+b) and non-treated (non-exposed, c+d). The theory behind it is mostly based on the fact that the (1-p) confidence interval for log OR is log(OR) +_ (plus or minus) z-value*SE where

SE = root of (1/a + 1/b + 1/c + 1/d). The entire mathematical steps are difficult to be written here, but the basic idea is to minimise the sum of squares via optimisation. 

Attaching a link to an article by Zhu Wang where both theoretical and practical implementation (in R) is shown.

Hope this was helpful for you 

 

Hi John.

Typically micro-social in application, yes. You marry this or that individual, or you deem this individual or this particular case eligible by some criterion, in this moment and this situation. (Although now you ask, I imagine we should, at least for the purpose of analysis, allow the possibility that there is a use of performativity higher up the chains of command, which might then apply to whole categories of individuals or cases.)

But the nature of a performative is that it entails attributing the characteristics of a pre-defined social/speech act category to any individual situation or case; and so it is the wielding, in practice, of a wider social construct. So there is the role of the meso- and the macro- in such constructing processes, of specific definition and of broader social significance, without which the particular instance would have no meaning.

What interests me is both the pervasiveness of such construction, in the gathering (the capta) of social data, and the extent to which such social process is acknowledged, in research or in the interpretations of policy makers.

I believe that attributin theory in psychoogy will be useful here, at least for the micro-level (as was ethnomethodology, in its brief heyday). But I woiud be interested to hear of any other areas of psychology, experimental or otherwise, that coid shed further light on this. What did you have in mind?.

My experience (like that of most people) is limited to a number of software packages. HLM is very easy to use. BUGS and GLLAMM (in Stata) are very flexible and cover the widest range of models, but both are challenged by large and complex data sets which can take a long time to estimate. MLwiN can handle large data sets very efficiently and can estimate models in likelihood and Bayesian (MCMC) mode- this allows goodish likelihood estimates to be used as starting values for the MCMC estimation. The MLwiN software also has lots of post estimation procedures to help interpret the results. MIXREG is very efficient and is very useful for discrete outcomes.

The Centre for Multilevel Modelling at the University of Bristol has recognised the problems with this and has started a large programme of work (sponsored by the UK ESRC) to provide for “inter-operability” – that is the ability to work across software platforms. The fruits of this so far are

1: runmlwin: Runs MLwiN from within Stata

2: R2MLwiN: runs MLwiN from within R

And much more ambitiously

3: Stat-JR

“is a brand new statistical software system: it constitutes a very different data analysis experience, featuring: an interface with a range of other statistical software packages, circumventing the need to learn software-specific techniques each time functionality of a new package required, but also providing tools to help teach software-specific knowledge to those wishing to learn; its own in-house MCMC based estimation engine (eStat) for complex data modelling (including multilevel models); open source templates allowing users to write their own Stat-JR functions; an eBook interface providing an interactive way of reporting and disseminating science, and an innovative tool for teaching statistics” Put simply you can specify a model in this environment and then ask for it to create syntax in a very large range of different software, as well as estimate with its own MCMC procedures .

This has just been released; see

Based on relationship of attributes Micro and macro are the domains you may categrorise but depends on the Association or ecological perspective, one to one relationship refers micro level and those of one to many or many to one/many may be Macro.

Dear Colleague, thank you very much for your answer which will help me and others to address the question in a well structured way. I am not a pessimist but my personal experience as a clinician is biased by the fact that I have seen some very unpleasant and unfortunate decisions in high risk and therefore complication-prone diabetes patients that have been driven by incentives such as private insurance coverage and networking based on financial criteria alone. I shall be pleased to discuss the subject in more depth in the future.

David Williams and colleagues developed a measure of Everyday Discrimination that is very useful at picking up on perceived poor treatment, slights, disrespect, etc, without initially identifying the reason for the discrimination. After those items, the individual is asked to report why they believe they were treated that way ( eg., race, age, gender, weight)

Causation in social epidemiology?

No, association does not indicate causality, either in social science or epidemiology; it only suggests where you might look, to find an underlying cause.

There was something of a revolution in sociology’s notions of causation in the mid-1960s. Until then, the dominant paradigm for causation had been the Durkheimian suggestion that we should “see social facts as things”, and in effect this meant treating social entities as external to the individual, and somehow causal in themselves.

In parallel with this – though originally developed as a counter to the Marxist notions of a dialectical materialist social dynamic – there was the school of thought that descends from Max Weber. This school of thought rejected the suggestion that we can explain social forms as caused at all, and instead looked simply to finding the meaning of social actions.

From the 60s, it would be fair to say that this more interpretative approach become the dominant approach, but with a new twist. Here, individual motivations, imbued with various cognitive processes, are what drives all interaction, and from that myriad of individual interactions, social processes and social structures are born. Nevertheless, individuals are not seen as isolated entities, but rather, as thoroughly and inescapably steeped in these socially constructed process of cognition and motivation, which surround us all. That is, social forms are both the consequence and, by impacting also on the cognitive processes, and the real world opportunities available - thus impacting on motivation – they are also external factors, and hence causal.

Once we can grasp this double-directionality in social process and structure, we will have no difficulty reconciling social facts and processes with medical ones, as epidemiology now does, since the same double-directionality applies also in epidemiology - and especially in mental health. So, for example, we can argue (and amass the statistics to corroborate the case, as Marmot, Wilkinson and Picket, Dorling etc do) that inequality impacts on health, via stress and disempowerment, and through motivating risk behaviour. But we can also see that inequality as itself being the product of social processes and mindsets.

There is simply no room here for the old Cartesian dualism, in which our bodies are physical things, and our social lives are immaterial. Our bodies constantly respond, both neurologically and hormonally, to experiences such as social events and interactions, including trauma. Meanwhile, our immune systems are designed (pace Dawkins) to shut down in fight/flight scenarios, and thus are disabled by chronic stress. And of course our limbs and guts are vulnerable to accidents and deliberate violence, pollution etc that are the result of social and economic structures.

So, social epidemiology looks for patterns in social life which echo or mirror patterns in health; and when we find these associations, we treat these as clues to follow, to look for the causal chains. From this evidence we develop hypotheses to test empirically. I could give examples, from my own field of housing studies; but then this posting will be FAR too long……

One example of "putting biological and social data together" has been studies on children and second-hand cigarette smoke in their household. Findings showed a greater prevalence of asthma, dental caries, and lower cognitive scores in school-aged children exposed to second-hand cigarette smoke. A biomarker for nicotine in their blood and hair was used in conjunction with parental self-reports of smoking behavior. Children, like pets, are unable to remove themselves from this exposure, which adds to the discussion of social environment, ethics, and biology.