| Copied from
http://www.deakin.edu.au/~agoodman/sci101/index.php
In this section you will look
at the ways in which science makes use of data, and how various
forms of data are also used outside science.
On completion of this section you should be able to:
- explain the importance of data collection, and give examples
of who collects data
- give two reasons why data may be collected without an
immediate and specific purpose
- list reasons why data (particularly in science) may be
collected for a specific purpose
- list the major ethical issues associated with undertaking
scientific research.
Contents
Introduction
Data Collection as an end in itself
Establishing the parameters of a system
Establishing benchmark data
Data Collection as part of broader strategy
Propaganda
Belief justification
Market research
Decision support
'Objective' research
The conduct of research
Introduction
How often have you heard someone say 'the fact is…' or 'the facts
speak for themselves'? We have an almost religious belief in the
importance of facts as immutable, independent, objective pieces of
information that tell us something 'real' about the world around us.
Our fascination with 'facts' is persistent and universal. They
seem to offer continual reassurance: whatever the foibles of human
opinion, some things at least are beyond argument. We all know that
up is up, left is left, and the sun rises in the east.
The people who are most consumed by the search for facts - at
least in the popular view - are scientists. To most people, science
is about the search for 'truth' - which is largely equated with the
accumulation of data. This magical material can be
organised into useful pieces (facts) from which laws can be
constructed. As the aim of science (so the argument goes) is to find
the 'laws of nature', everything the scientist measures is data, and
every piece of data is potentially important. In its extreme form,
this approach sees science as the process of collecting (and
sifting, organising and summarising) masses of data. In this
scenario data have particular and special significance.
The problem with this view is that it is almost totally wrong. It
is true that some scientists collect data (usually resulting from
experiments) but many never handle data in the conventional sense.
Few, if any, scientists see the accumulation of data as worthwhile
in its own right. Most scientists know that data are only useful in
the right context; data out of context are at best unhelpful, at
worst misleading. Good scientists in particular have an instinct for
knowing which data are useful and relevant, and which are not.
Nor does science progress (if indeed science can be said to
progress) by the mere accumulation of facts. The popular image of
the scientist has not caught up with modern thinking about how
science is conducted. Society's conception of scientific method
is quite different from that accepted as appropriate (and
preferable) by those who study the process of science. In their
analysis there is a right way to do science and a wrong way - and
the science of popular conception is the wrong way. When you
understand the distinction between these approaches, you will see
more clearly what role data play in these methods.
Of course, scientists in the traditional fields (physics,
chemistry, biology, astronomy and so on) are not the only people who
collect data. The modern world - complex, industrialised,
bureaucratic - thrives on data of all kinds. We are numbered,
analysed and surveyed throughout our lives, and the results are
stored and analysed. We have in some senses become a part of the
statistics that largely define modern society.
But who actually collects data? All governments do, for reasons
both laudable and questionable. Without up-to-date and comprehensive
data about the characteristics of the population no government can
plan and build the facilities and resources we have come to expect.
Commercial organisations collect data to improve their economic
prospects by offering the goods or services that potential customers
seem to want. Researchers collect data to further their
understanding of the workings of our social and economic systems.
Physical scientists collect data to further their understanding of
how the world functions.
The process of collecting data takes two forms: gathering data
that already been collected by someone else (probably for a
different purpose), and creating 'new' data. The latter is a matter
of some philosophic importance, and we will also return to it
shortly.
Data Collection as an end in itself
What motivates people to go through the often complex and costly
process of collecting data? Apart from simply collecting information
to satisfy a fascination with so-called trivia, two main reasons for
collecting data without an immediate and specific purpose are:
- to establish the parameters of a system
- to establish benchmark data
Establishing the parameters of a system
When we investigate natural and social systems we often start
with no clear idea of how the system functions. In particular, we
may have no strong impression of how the properties of the
components of the system may vary. If we are studying river flows,
for example, we may have no idea of the likely range of values to be
expected in a particular system. We can get some idea from studies
of rivers in similar environments, but this may not really be
transferable due to some peculiarity in this location. In this case
we need to carry out preliminary studies whose results will define
the parameters of the system. These parameters will be
concerned with the probable extremes of the data we expect to find
in the 'real' study, and the likely variability of the data. This
knowledge may have a direct impact on the way in which we collect
data during the major part of the study.
Establishing benchmark data
If we pursue the example of river flow studies, we can also
illustrate the way in which data are sometimes collected to
establish benchmarks. Regular monitoring of river levels,
even if this is not part of a specific study, will help to build a
picture of the general behaviour of the system. This will provide
valuable comparisons and context when we study the system in more
detail. Establishing benchmark data on flow patterns will indicate
how 'typical' the data collected are for a particular time period;
they will also reveal long-term changes in the system.
Data Collection as part of broader strategy
Most data are collected for a specific purpose, as part of a
broader strategy. We may be surveying how people would react to a
political proposition, or the likely sales for a new product. We may
be investigating the effect of airborne pollution on vegetation
systems, or measuring the mass of a newly-discovered atomic
particle. The following are some of the main reasons why people
collect data:
Propaganda
Some data are collected for what we might call propaganda
purposes: to convince other people of the rightness of your view, or
a group to which you belong. Most propaganda that involves real data
is based on processing and presenting raw data in a way that suits a
particular message, rather than on the generation of new data.
This category could also include instances of scientific fraud in
which data are falsified or misrepresented to convince a scientist's
peer group of the correctness of his or her work.
Belief justification
Many people seek data that support the views that they already
hold; this is true of some scientists, although it is rare.
Market research
Enormous amounts of data are collected by commercial
organisations about the buying intentions of consumers; these
surveys are also widely used in the social and political arenas.
Decision support
In industry and government we have become used to the expectation
that decisions will be based on careful analysis of data. For
example, before building a new port, an environmental impact
statement would be carried out. Relevant data about the area to be
affected are collected and collated; this would then form one of the
key foundations of the decision about whether or not to proceed with
that development.
"Objective" research
Scientists (and those who aspire to that status) collect data as
a critical part of the process of research. You'll see how this
process operates in detail - and its implications for the data
collection process - in later sections. You'll also examine the
popular (but misleading) concept of objectivity.
Exercise 1.1
Look through a number of issues of different magazines and
newspapers for examples of data being used for specific purposes.
Try to classify them using the categories defined in this section.
Ethical issues
Unlike most experimenters in physics and chemistry, or many in
the earth sciences or biology, who deal with inanimate objects or
materials, researchers whose subjects are people or animals must
consider the conduct of their research, and give attention
to the ethical issues associated with carrying out their research.
Sometimes a research project may involve changing the subjects'
behaviour or, in some cases, causing them pain or distress. Most
research organisations have complex rules on human and animal
experimentation. Although a detailed study of such rules is beyond
the scope of this unit, you should be aware that such systems exist,
and what they deal with.
The American Psychological Society has developed a set of
guidelines governing the conduct of research in psychology. Although
some are clearly most relevant to psychology, most are applicable to
all forms of research, and will give you an impression of the
ethical issues involved. They can be summarised as follows:
You must justify the research via an analysis of the balance
of costs.
The scientist's interest alone isn't sufficient justification to
carry out research. In order to carry out experiments there have to
be benefits that outweigh the costs. Researchers are expected to
carry out an analysis and ensure that the research is justified.
You are responsible for your own work, and for your
contribution to the whole project.
Scientists must accept individual responsibility for the conduct of
the research and, as far as foreseeable, the consequences of that
research.
You must obtain informed consent from any subjects.
The concept of informed consent is a major problem when
dealing with research into human behaviour or physiology,
particularly in research that may have harmful side-effects. Can
someone give informed consent, for example, if they are below 18
years of age (or whatever is the legal age of consent)? What about
potential subjects who are mentally or physically disabled?
You must ensure that all subjects participate voluntarily.
In psychological research all subjects must participate voluntarily;
informed consent must be accompanied by a free decision to
participate. On the other hand, it is very difficult to explain
complicated research to non-specialists. Nevertheless, the onus is
on the researcher to explain the research, not on the volunteer to
find out about it.
You must be open and honest in dealing with other researchers
and research subjects.
The researcher must be as open and honest as is reasonable. This
process, whilst fine in principle, is complicated by considerations
such as commercial advantage and professional rivalry.
You must not exploit subjects by changing agreements made
with them.
As a researcher you might discover that your experiment shows
something that you would like to further investigate, but don't want
to tell your subjects about. If you did investigate further, but
pretended that you were still doing the experiment that had been
agreed to in the first place, this would be a form of exploitation,
and would breach the principles of informed consent and voluntary
participation.
You must take all reasonable measures to protect subjects
physically and psychologically.
The unexpected outcomes of a series of now-famous experiments in the
1950s convinced most psychologists that even voluntary participants
can 'get carried away' to the point where they have to be protected
from themselves and each other. In these experiments, university
psychology students were allowed, using complex behavioural rules,
to 'punish' their co-subjects where they breached the rules. What
surprised and disturbed the researchers was the number of students
who greatly exceeded their 'right' to inflict punishment, and how
widespread the process became among the subjects. Eventually the
experiments had to be terminated to prevent injury to the subjects.
The researcher must be prepared to intervene, even at the cost of
the experiment itself, to protect the subjects.
You must fully explain the research in advance, and
'de-brief' subjects afterwards.
Whilst full explanations before the experiment are essential to
gaining informed consent, it is, unfortunately, a common practice
for researchers to complete their research without telling the
participants anything about the results.
You must give particular weight to possible long-term effects
of the research.
Obviously this can be difficult to achieve. It means that,
regardless of their strong motives for collecting information,
researchers have to give particular emphasis to any potential
long-term dangers of the research.
You must maintain confidentiality at all times.
Only certain people conducting the experiment should know the
identify of the participants, and any subject should generally not
know the identity of other subjects. The key to maintaining
confidentiality is that the individual should not be identifiable.
|
