IB Biology HLPractical SkillsPaper 1B & IA~13 min read
Using Tech to Process Data
A spreadsheet will happily calculate a mean to nine decimal places from three readings, and draw you a beautiful chart of something meaningless. The technology removes the arithmetic, not the thinking — and the marks are for the thinking.
📚 What you need to know
The volume and complexity of data from some investigations makes technology essential for processing, analysing and interpreting it.
Spreadsheets record and manipulate data: raw data is input, categorised, and organised into columns and rows.
Spreadsheets perform calculations, statistical analyses and mathematical operations on datasets.
Computers draw graphs from raw data, using built-in functions to visualise trends, patterns and correlations.
Data with large ranges, such as population data, is much easier to manipulate on a computer.
Computers use data to produce models that inform ongoing predictions.
Images can be analysed using computer programmes — for example images of joints in motion.
Spreadsheets
A spreadsheet does three separable things, and it is worth naming them separately because questions tend to target one at a time.
Recording, calculating and plotting are three different claims to make about your data. Questions asking “how did technology help?” usually want you to name which of the three, and why it mattered here.
Recording
Raw data is easy to input, categorise, and organise into columns and rows.
The convention is one row per observation and one column per variable, which is what lets the software treat a column as a set.
Keep raw data untouched, and put processed values in new columns.
Calculating
Spreadsheets perform calculations, statistical analyses and mathematical operations on whole datasets at once. Where a calculator handles one number at a time, a spreadsheet applies the same operation down a column of two thousand.
The ones you will actually use: mean of repeats, standard deviation to describe spread, percentage change, the gradient of a line to get a rate, and correlation between two variables.
Plotting
Built-in functions generate graphs and charts automatically, making it possible to visualise trends, patterns and correlations. This is particularly valuable for data with a large range — population data being the classic example, where numbers may run from tens to millions and a computer can rescale the axis instantly.
The default chart a spreadsheet offers you is almost never the right one. It will cheerfully join categorical bars with a line, start a y-axis at a value that exaggerates a tiny difference, or add a trend line through data that has no trend. Choosing the chart is a biology decision, not a software one.
The last line is the reason bar charts have gaps. Joining “habitat A” to “habitat B” with a line would suggest there is something halfway between them, and there is not.
Producing models
Computers use collected data to produce models that inform ongoing predictions. A model built from real measurements is fitted to the data, then used to estimate values you did not measure.
Two words are worth separating here:
Interpolation — estimating a value within the range you measured. Usually reasonable.
Extrapolation — estimating a value beyond that range. Risky, because you have no evidence that the pattern continues.
The enzyme practical makes this vivid. Rate rises steadily with temperature from 10 °C to 40 °C, so a fitted straight line will confidently predict a very high rate at 70 °C. Biology says the enzyme will have denatured long before that. The model does not know about tertiary structure.
Image analysis
Images can be analysed using computer programmes. The example given in the course is images of joints in motion: video of a moving limb is broken into frames, points on the joint are marked, and software calculates the angle at the joint in each frame.
Doing this by hand with a protractor on printed stills is possible but slow and imprecise. Software gives a value for every frame, so you get angle against time — and from that, the speed at which the joint moves.
What technology does
What it cannot do
Calculates a mean from 2000 values instantly
Tell you whether those values were valid to collect
Draws a line of best fit through any data
Know whether a linear relationship is biologically plausible
Extends a model beyond the measured range
Know that the enzyme denatures at 60 °C
Reports a correlation coefficient
Establish that one variable causes the other
Removes arithmetic slips
Remove a systematic error in the raw readings
The line examiners reward. Correlation is not causation, and software cannot tell the difference. A spreadsheet will report a strong positive correlation between ice cream sales and drowning incidents. The biology — a third variable, hot weather — has to come from you.
Worked examples
WE 1
Choosing and justifying a graph
A student measures the mean number of limpets per quadrat at five fixed distances up a rocky shore. State which type of graph should be used to display the results, and justify your choice. (3 marks)
Step 1: identify the variables
Distance up the shore is continuous, and it was chosen by the student, so it is the independent variable on the x-axis. Mean number of limpets is the dependent variable.
Step 2: state the graph
A line graph, with distance on the x-axis and mean number per quadrat on the y-axis.
Step 3: justify it
Because distance is continuous, intermediate values exist, so joining the points is meaningful. A bar chart would wrongly imply the five distances are separate categories.
Line graph, because the independent variable is continuouserror bars showing the spread at each distance would strengthen this considerably.
WE 2
Reading a spreadsheet’s statistics
A spreadsheet reports the mean height of plants in two conditions. Condition A: mean 24.6 cm, standard deviation 3.2 cm. Condition B: mean 27.1 cm, standard deviation 2.9 cm. A student concludes that condition B causes taller growth. Evaluate this conclusion. (4 marks)
Step 1: what the means show
The mean in B is 2.5 cm higher, so there is a difference in the data as collected.
Step 2: what the spread shows
Using one standard deviation, A spans roughly 21.4 to 27.8 cm and B spans 24.2 to 30.0 cm. These ranges overlap substantially.
Step 3: the conclusion this allows
The overlap means the difference between the means may be due to random variation rather than the condition. A statistical test, such as a t-test, would be needed before claiming a real difference.
Step 4: the word “causes”
Even a significant difference would show association. Causation requires that all other variables were controlled.
The conclusion is not supported — the spreads overlap too mucha spreadsheet will report a difference between any two means. Whether it means anything is your judgement, not the software’s.
WE 3
Percentage change, and the limits of a model
A population is recorded as 1200 individuals in 2015 and 1860 in 2020. Calculate the percentage increase. A model fitted to this data predicts 4400 individuals by 2035. Suggest why this prediction should be treated with caution. (4 marks)
Step 1: find the increase1860 − 1200 = 660 individualsStep 2: express it as a percentage of the original(660 ÷ 1200) × 100 = 55 %Step 3: why the prediction is uncertain
It is an extrapolation well beyond the measured range, from only two data points, and it assumes growth continues at the same rate.
Step 4: the biology
Real populations are limited by a carrying capacity — food, space, disease and predation increase as numbers rise — so growth is likely to slow and the model will overestimate.
A 55 % increase, but the 2035 figure is an extrapolationalways divide by the original value for percentage change. Dividing by 1860 gives 35 % and loses the mark.
💡 Exam tips
Name which of the three jobs technology did: recording, calculating or plotting.
Justify a graph type from the independent variable: categoric means bars, continuous and chosen means a line, both measured means a scatter.
Percentage change divides by the original value.
Use interpolation and extrapolation by name when discussing predictions.
When two means differ, check whether the spreads overlap before claiming a real difference.
Say correlation does not establish causation whenever a question invites you to leap from one to the other.
⚠ Common mistakes
Accepting the default chart. The software picks a shape, not a meaning.
Joining bars with a line. Separate categories have nothing in between them.
Quoting a mean to more decimal places than the raw data. A spreadsheet will offer nine; your instrument earned two.
Extrapolating a trend line off the end of the data. Biology usually stops the trend before the maths does.
Overwriting raw data with processed values. Once it is gone you cannot check anything.
Treating a strong correlation as proof of cause. It never is on its own.
Pulling the skill set together
Two pages, one arc. Technology lets you collect data you could not otherwise gather — faster, for longer, at intervals no person could keep up with — and then handle volumes of it that would take weeks by hand.
What it does not do is decide whether the data was worth collecting, whether the sensor was calibrated, whether the graph is honest, or whether the pattern means what it appears to mean. Those judgements stay with you, and they are where the marks live.
In your internal assessment, the strongest sentences about technology are specific ones. Not “a data logger was used to improve accuracy”, but “readings were logged every 2 s because a preliminary trial showed the reaction was over within 30 s, and manual timing would have produced only three data points.”
Want this explained one-to-one?
Book a free session with an experienced IB Biology tutor and get your trickiest topics made simple.
