File : THESIS.pdf - Universitas Muhammadiyah Malang

assumptions. Trend lines should show an even distribution of points on
either side of the line along its whole length. Lines should be finely drawn
and should not contain kinks or breaks.
Analysis, conclusions and evaluation
Interpretation of data or observations and identifying sources of error
Candidates should be able to:
• describe the patterns and trends shown by tables and graphs;
• describeand summarise the key points of a set of observations;
• find an unknown value by using co-ordinates or axis intercepts on a graph;
calculate other quantities from data or from quantitative data related to their
qualitative observations, or calculate the mean from replicate values, or
make other appropriate calculations;
• determine the gradient of a straight-line graph or tangent to a curve.
• evaluate the effectiveness of control of variables and thus the
confidence with which conclusions might be drawn;
• identify the most significant sources of error in an experiment;
• estimate, quantitatively, the uncertainty in quantitative measurements;
• express such uncertainty in a measurement as an actual or percentage error;
• show an understanding of the distinction between systematic errors and
random errors. Descriptions should be precise, giving quotations of figures to
support the description, and calculated values where these are appropriate.
Unknown values might include unknown concentrations where a calibration
curve has been drawn, or values for 50% plasmolysis or zero change in mass
in osmosis experiments. Calculations may involve mean, percentage,
percentage gain or loss, rate of reaction, magnification, actual size or other
appropriate calculations. When a gradient is to be determined, the points on
the line chosen for the calculation should be separated by at least half of the
length of the line or tangent drawn. Candidates should be used to looking at
experiments and assessing the relative importance of errors in measurementor
in making observations so that they can judge which sources of error are most
important. Candidates should be familiar with simple means of estimating
error, such as the errors intrinsic in measuring devices (see
http://www.chemistryreact.org/go/Tutorial/Tutorial_4428.html)
or in the observer’s ability to
observe, or in experiments where limitations of the method introduce errors
(e.g. heat loss when trying to assess the energy content of biological
materials). They should be able to express these errors in standard forms
such as length = 73 mm ± 1mm, or temperature increase = 14 ˚C ±
4 ˚C. Candidates should be able to suggest which of the sources of error
described are likely to be systematic errors such as those resulting from
thermometers that consistently read 1 ˚C above actual temperature, or
candidates who read volumes to the wrong part of the meniscus, as well as
those which are likely to be random errors due to variability of biological
materials, or random variations in room temperature.
For key control variables, candidates should be able to give a realistic
estimate or appraisal of how effectively the variable was controlled, for
example, how closely the temperature was maintained the same across a
number of samples, and from this, give an indication of the confidence that
they would have in any conclusions drawn.
Drawing conclusions
Candidates should be able to:
• draw conclusions from an experiment, giving an outline description of the
main features of the data, considering whether experimental data supports
a given hypothesis, and making further predictions;
• draw conclusions from interpretations of observations, data and calculated
values;
127