APECS Workshop

... on field based research methods and project design.

APECS Sweden organized a workshop two weeks ago here in Stockholm about field work in polar regions. I hope that I will be able to participate in an expedition to the polar regions sometime in the future. But no matter what, it is interesting to listen to other's experiences and advice - all the way "From Idea to Publication".

One thing that I thought about a lot was the scientific method. The classical way is to go from a hypothesis or theory about something and then design an experiment based on what you predict or deduce, to test your idea. But today, many times it's the other way around. You have an experiment, and from that you try to find a hypothesis and test it. Which way is the best?

Anyway. A successful field campaign depends on a lot of things. First, you need to formulate your research question and also ask yourself what you need in order to investigate that.

1) Formulate your research question
2) Answer the questions:

• what do I need?
• what data can I feasibly gather?
• what limitations do my data have?
• which (if any) statistical test should I use?

3) Practice at home and test run everything.
4) Talk to experienced people.
5) If possible, plan your campaign to answer several hypothesis.

These are all good advice, I think. I wonder if all research project leaders have this approach? I'm sure there are several method on how to do this, but the result might be the same.

We also saw some presentations with more technical information. Thing that you might think about every day in front of your computer, but you never talk about them. For example, different data types:

• Nominal/categorical - descriptive (large, small, warm, cold)
• Ordinal/Rank - comparable (median, percentile)
• Interval/Scale - ex. temperature (mean, std, corr)
• Ratio - mass, length, time

There was also a discussion about Pseudo-replication, when individual replicates within a group are not independent. It's often unavoidable in practical research. But we should not pool replicates (temporal or spacial) prior to analysis.

Another thing to think about when doing field research is the scale. All spacial studies need to choose an appropriate scale and try to keep all the input variables at the same scale.

We also need to avoid Spacial Autocorrelation, where sites close to each other will be more alike than sites further away from each other. If you don't know anything about an area it is probably best to make a random distribution of your measurements. Remember, there is a difference between accuracy and precision.

A useful presentation was about logistics, equipment and safety in the field. This reminded me a lot about all the things that we had to think about during the course on Svalbard, earlier this year. In polar research, you can't really separate between science and logistics, because everything is dependent on each other.

First, you have your personal goals, then you have your scientific goals and equipment. On the other side you have the cold and harsh nature. Between the science and the nature, you have logistics and that can be divided into three things; comfort, environment and safety.

The comfort is usually your responsibility and it includes transportation, sleep, eat, toilette & hygiene, communication, personal equipment, social & free time. Tips and tricks are;

• plan one box at a time
• think is systems
• don't forget spares and repairs
• something falling between the boxes?
• check for duplicates

There are a lot of things to think about when you talk about safety:

• 1st Aid equipment
• Safety equipment for the area - Avalanches, sea ice, cold weather
• Training/Education for handle field environment e.g. glaciers
• Ice information & trip plans
• Protection - polar bears - rifle, signal gun, dogs
• Spares and repaires  

The scientific planning is also very long and it is important that everyone know what to do when you are out on the field. Some things to think about when doing this;

• Plan for 120% field work
• Define 100% and enjoy the bonus
• Do your priorities early and discuss them!
• If there are several projects going on - discuss the differences
• Don't be disappointed with 30%...
• Plan enough time
• Have a group leader and a trip plan
• Functional test of equipment - computability with other things?
• Don't forget the emotional stress!
• Try to keep daily or weekly routines
• Take care of each other
• The difference between good & bad management will have an influence on the group energy!

More key points and general tips:

• Field work is time consuming
• Structure your work
• Plan is systems
• Delegate the work
• Rank your asssistants
• Prepare for the worst
• Don't forget small issues
• Prepare yourself for physical work!
• HAVE FUN!

Before going:

• Bring spares and repairs (again!)
• Start packing in field-ready containers as early as possible
• Don't forget office supplies
• Don't forget normal lab-equipment (plastic bags, markers, etc..)
• Pack personal belongings
• Notify people that you will be away!
• Write checklists! Numbers on a items.
• Safety equipment before and after the expedition?

This is a very good suggestion (from my experience working with field data, sometimes you wonder what happened on a specific day): Write down as much as possible during the expedition! Try to finish the cruise report already on the ship. Keep a research diary! Also very good to do in your office...

The last part was about Data Processing - what do you do with all the data that you've now got? Well, it's really important to organize it! Instead of having many different directories named "new", "newest",  "old" or "final version 1", final version 2" and so on... there are better ways!

There are two things that you can organize your data after: source or project. For example;
data-o: original data
data-der: derived data using general conversions (format change, average...)
data-res: research results data, for specific questions or projects

This was very useful for me because my directories can be quite messy sometimes. However, I do have the routine to sort them out every now and then... but it would of course be better if everything was nicely sorted from the beginning.

Some key notes here are:

• script everything - it is likely that you need to do things severals times
• choose tools wisely - the chose is likely to become permanent
• think about open source options for data processing

It was a really nice workshop at the Royal Academy and it was also nice to see so many young polar scientists from all over Sweden. I hope people will keep coming to the Swedish APECS meetings.

the most important point that I will remember from this workshop: 

organize your data and keep a research diary.