><> ><> ><> ><> ><> ><> ><> ><> ><> ><> ><> ><> ><>
REPLIES WILL BE SENT TO THE FISH-SCI LIST
<>< <>< <>< <>< <>< <>< <>< <>< <>< <>< <>< <>< <><
An interesting challenge!
Any kind of statistical analysis would require you to determine which
units (individual vessel days, perhaps?) are replicates of other such
units. ANOVA would require that you bin the data in some way. Do not
rush those steps. You may learn as much about the fisheries from
thinking about how to arrange the data as from the analyses that
follow. Indeed, the analyses could all too easily return results that
are merely artifacts of the way that the data were arranged and hence
getting the right arrangement could be critical.
Twenty years ago, I worked on a data set that was a lot smaller (some
200 boats) but more detailed (being based on interview data) than
yours likely is. One thing that quickly became evident was that the
200 captains followed about 150 different annual cycles -- quite
deliberately so, as (outside of the local lobster seasons, which
encouraged everyone to do much the same as his neighbours for several
weeks) most men consciously chose to spread themselves across the
available fishing opportunities by specializing differently from
everyone else. The concept of a "replicate" became very shaky.
I'd recommend a lot of exploratory, descriptive analysis, before
moving to hard number crunching.
Also: Fish catches are a consequence of fishing effort (as well as
the availability of fish biomass to harvest) and fishing effort is,
ultimately, a product of human decisions. As a sometime biologist, I
find it all too easy to analyze catch data as though it was the
outcome of some sort of bio-economic process but, in reality, it is
mostly a result of the choices of the most behaviourally-complex of
all species: Homo sapiens. The one saving grace is that we can talk
to the decision-makers and so come to an understanding of their
choices without the sort of manipulative experiments needed when
working with other species -- experiments that would be both too
expensive (because of the behavioural complexity) and unethical if
attempted with fishermen. In short and if you have not already done
so: Discuss your data and your analyses with the subjects of your
study. They may well help you define the questions to be asked of the
data, as well as aiding interpretation of your analytical results.
Trevor Kenchington
On 3-Jun-14, at 2:30 PM, Grant Adams wrote:
>> <> ><> ><> ><> ><> ><> ><> ><> ><> ><> ><> ><> ><>
> REPLIES WILL BE SENT TO THE FISH-SCI LIST
> <>< <>< <>< <>< <>< <>< <>< <>< <>< <>< <>< <>< <><
>
> I am assisting a South American research institute. After working on a
> commercial species catalog and noticing the lack of published local
> fisheries data
> I have convinced my colleagues to publish. We are going to start
> with a 10 year
> data set of daily artisanal landings (kg) including; date, vessel,
> gear, species,
> and fishing grounds.
>
> Currently we are thinking of describing the seasonal and
> interannual/long term
> trends of all pelagic and coastal artisanal fishery operations
> using ANOVA and
> multivariate regression.
>
> We are looking for the best way to diffuse this information;
> analysis, focus, etc.
>
> Any advice would be great!
>
>> <> ><> ><> ><> ><> ><> ><> ><> ><> ><> ><> ><> ><>
> For information, send INFO FISH-SCI to [log in to unmask]
>
> The FISH-SCI List Archive
> http://segate.sunet.se/cgi-bin/wa?A0=FISH-SCI
>
> To cancel your subscription, send a blank message to:
> [log in to unmask]
> <>< <>< <>< <>< <>< <>< <>< <>< <>< <>< <>< <>< <><
><> ><> ><> ><> ><> ><> ><> ><> ><> ><> ><> ><> ><>
For information, send INFO FISH-SCI to [log in to unmask]
The FISH-SCI List Archive
http://segate.sunet.se/cgi-bin/wa?A0=FISH-SCI
To cancel your subscription, send a blank message to:
[log in to unmask]
<>< <>< <>< <>< <>< <>< <>< <>< <>< <>< <>< <>< <><
|
