NADRA 2018: Challenges faced by researchers
26 sep 2018
This year the key note presentations at the Nordic Alcohol and Drug Researchers’ Assembly (NADRA) discussed different challenges researchers tackle in their work.
Senior researcher Pia Mäkelä from the National Institute of Health and Welfare in Finland talked about the interplay between alcohol researchers and policy makers and specifically about the importance of reaching out with your research results.
– Researchers feel comfortable doing the actual research but when it comes to disseminating and communicating our results, we´re usually outside our comfort zones. The end point of a research project should, however, not be ”one more paper in the library” but making an impact or, at a very minimum, ensuring that the people who care about and need our results hear about them, Mäkelä says.
She states that in order to reach decision makers these days, the researchers have to reach out to them in various ways: via news releases, web pages, blogs, social media, popularizing journals, responding to media enquiries etc.
– For some rare research institutions with good resources communication experts take care of the communication and popularization, but most of us have to be prepared to do a lot of the work ourselves, says Mäkelä.
A balancing act
The shift from production of information to dissemination and support in policy processes is not always easy for a researcher.
– If researchers wish to promote population health and well-being, they should take a stand and try to influence policy, i.e. to have opinions in addition to producing information. But, once you have an opinion as a researcher, there is a risk that you are no longer seen as the objective researcher. It`s not an easy balancing act, says Mäkelä.
She gives some advice on how to communicate, on the basis of her experience from the process leading to the new Finnish Alcohol Act.
– Do share information and respond to questions, even if they don´t concern your own research findings but the field more generally. Be aware of “falce balance” interviews; researchers versus lobbyist may sometimes be a doomed design. Identify central actors in the process of preparation of legislation. The timing of the policy process is also of importance, don´t miss the windows of opportunity in the legislative cycle to make an impact, says Mäkelä.
How to gain trust as an ethnographer
Professor Philip Lalander from Malmö University was the second key note speaker at NADRA 2018. His presentation focused on his long-term ethnographic project on heroin users and dealers in Norrköping (from 2000 to 2016) and especially on how to gain trust as a listener and as a person of meaning.
Lalander says that it is very important to have the ability to create an impression of a person who will do something valuable with the information obtained from interviews and observations.
– For me, it played a significant role that the informants felt I was committed and doing something for them – and that I would write a book about the project.
Lalander says that in order to create the possibility for a sustainable contact and relationship, you have to become a person of meaning, which gives you a big responsibility.
– I did not just want them to talk to me once. I wanted to both talk to them and spend time in everyday life with them as my guides.
Interested, not curious
According to Lalander, the maintaining of access has to do with how motivated the informants are to talk about their lives. The researcher can thus be seen as a motivator who can influence the possible interaction in an encounter.
– If a participant feels unmotivated or distrusts the researcher he or she may not answer the phone, fail to show up or bad-mouth the researchers, reinforcing a bad reputation.
Lalander´s hands on-advice is to meet the informants alone because it shows that you don´t see them as stigmatized people to be afraid of. It´s also important to show interest and be able to small-talk about everyday stuff, like sports or movies.
Lalanders says that the informant´s motivation increases when they feel they are contributing to the research project.
– See the persons as fellow human beings who have experience and knowledge to convey. It also helps if they feel their participation could lead to something good; to counteract simplistic stereotypes for example. And it helps if the researcher is interested, eager to learn, rather than being curious.
Declining participation levels in surveys
Health surveys and the challenge of declining participation levels was the theme of the third key note presentation at NADRA 2018. The presentation was given by Senior investigator scientist Linsay Gray from University of Glasgow.
Gray talked about how the continuing decline in participation in population-based health surveys has an effect, not only on the power of analysis but, crucially, on the reliability of inference.
– The consequence of non-participation in health surveys is potential for bias with under-represented groups, such as lower socio-economic groups, males, younger age groups, singles, those in poorer health and those with higher levels of risky behaviours, Gray explains.
There are, however, statistical methods available that can be applied to help correct for non-participation bias. With advances in methods for dealing with non-participation, the aspiration is for surveys to continue to be conducted as sources of reliable information on population health, says Gray.
Potential solutions Gray talked about combine record linkage, multiple imputation and sensitivity analysis.
Record linkage connects records across different data sources – such as survey and administrative health data – that refer to the same individual. Multiple imputation is a general approach to missing data, that allows for uncertainty by creating different imputed data sets from the observed data and combining their results.
– The biggest challenge is dealing with non-participation that is a consequence of the survey variable of interest. This can be done via a type of sensitivity analysis in multiple imputation known as pattern-mixture modelling which involves changing the imputation model to reflect potential differences in the distribution of the estimate between participants and non-participants, incorporating information from external sources, Gray explains
Gray concluded her presentation with some good news for the Nordic researchers.
– As the register-based systems in operation are amenable to record linkage and direct identification of individual non-participants, the Nordic countries are well placed to implement these methods.
21 jan 2021
7 mai 2020