IB Biology SLSkill Set 2 — Using TechnologyPaper 1 & 2Practical skill~10 min read
Using Tech to Process Data
Collecting data is the easy half. Once you have 600 readings from a logger, or a table of repeats from a whole class, you need to turn them into something a reader can understand — a mean, a graph, a pattern. A spreadsheet will do all of that in seconds. It will also do it wrong, without complaining, if you point it at the wrong cells.
📘 What you need to know
Spreadsheets store raw data in rows and columns, then calculate, sort, filter and analyse it.
A formula refers to cells rather than numbers, so the answer updates itself when the data changes.
Computers plot graphs automatically — but choosing the right type of graph is still your job.
Error bars show the spread of your repeats. Overlapping error bars mean the difference might be chance.
Data with a huge range is easier to handle on a logarithmic scale.
Computer models use your data to make ongoing predictions.
Image analysis software measures and counts from photographs and video — joints in motion, colonies on a plate, cells on a slide.
Spreadsheets: rows, columns and formulas
A spreadsheet is a grid. Each column holds one variable, each row holds one set of readings, and every cell has an address made of its column letter and row number — B2, C7, and so on.
The powerful bit is that a formula points at addresses, not at numbers. Type the mean of B2 and C2 as a formula, and if you later fix a typo in B2, the mean corrects itself. Do the same sum on a calculator and you have to redo every single one.
Notice the rate peaks at 40 °C and then falls — the enzyme is denaturing. A spreadsheet finds that shape for you in one click, once the data is laid out properly.
Formulas worth knowing
What you want
Formula
What it tells you
Mean of the repeats
=AVERAGE(B2:C2)
The typical value for that condition
Standard deviation
=STDEV.S(B2:F2)
How spread out the repeats are
Range
=MAX(B2:F2)-MIN(B2:F2)
The gap between highest and lowest
Total
=SUM(B2:B6)
Adds a column or a row
How many readings
=COUNT(B2:B6)
Useful for checking nothing is missing
Percentage change
=(C2-B2)/B2*100
Change relative to the starting value
WORKED EXAMPLE
Five repeats of a reaction rate give 24.6, 26.1, 25.4, 27.2 and 25.7 arbitrary units. Find the mean and the standard deviation, and say what each one tells you.
Step 1: the mean=AVERAGE(B2:F2) gives 25.8Step 2: the standard deviation=STDEV.S(B2:F2) gives 0.96Step 3: what they mean
The mean is the value you plot. The standard deviation says the repeats sit within about one unit of it, so this is a tight, repeatable set.
Mean 25.8, standard deviation 0.96A mean on its own hides everything. Always report the spread alongside it.
Letting the computer draw the graph — but choosing it yourself
Spreadsheets will plot anything you ask for, including nonsense. The software has no idea whether your independent variable is categoric or continuous, so this decision stays with you.
A line of best fit is not a dot-to-dot. It is a single straight or smooth line showing the trend, with points scattered either side of it.
WORKED EXAMPLE
A student measures bubbles produced per minute by pondweed at six different light intensities. Which graph should they plot, and what goes on each axis?
Step 1: what kind of variable is the independent one?
Light intensity is continuous — it can take any value in between.
Step 2: so which graph?
A scatter graph with a line of best fit. Bars would wrongly suggest the six intensities are separate categories.
Step 3: the axes
Light intensity on x (independent), bubbles per minute on y (dependent), both with units.
Scatter graph with a line of best fitIf you can ask “what happens halfway between these two values?”, the variable is continuous.
Error bars: how much do you trust the difference?
Once you have a mean, plot the spread with it. Error bars usually show the range or the standard deviation, and they turn a bare bar chart into something you can actually argue from.
Overlap is a warning, not a verdict. It tells you the difference could be down to variation between repeats rather than the variable you changed.
The sentence examiners want is something like: “the error bars overlap, so the difference between the two means may be due to chance rather than the treatment.” Learn the shape of it — it fits almost any data question.
When the numbers span a huge range
Population data is the usual culprit. If your counts run from 100 to 1 000 000, a normal axis squashes everything small into a flat line against the bottom.
A logarithmic scale fixes this: each step up the axis multiplies by ten, so 100 and 1 000 000 both fit and both stay readable.
On a log scale, exponential growth appears as a straight line, which makes it obvious at a glance.
The trade-off is that log axes are easy to misread. A small-looking gap can be a ten-fold difference, so label the axis clearly.
Models and predictions
Once your data is in a spreadsheet, the computer can fit a relationship to it and use that to predict values you never measured.
Fitting a line or curve to your points gives you an equation for the trend.
Interpolation — reading a value from between your data points — is usually safe.
Extrapolation — extending beyond your data — is a guess. Biological systems have limits: enzymes denature, populations hit a carrying capacity, and the neat line stops being true.
Models are only as good as the data and assumptions you feed them.
Careful with trend lines: a spreadsheet will happily fit a straight line to a curve and report how well it fits. That number describes the fit, not the biology. Look at the shape of the points yourself.
Image analysis
What it is for: measuring and counting from a picture instead of from the specimen. Software is faster than you, does not get bored, and gives the same answer twice.
Movement: video of a joint in motion can be analysed frame by frame to measure the angle at the elbow or knee as it bends.
Counting: software counts bacterial colonies on an agar plate or cells in a microscope field, far faster than clicking a counter.
Measuring: leaf area, root length, stomatal density, cell diameter — all measurable straight from a photograph.
It all rests on the scale. A photograph has no size of its own, so a scale bar or a known reference length has to be in the image.
WORKED EXAMPLE
On a micrograph, a scale bar labelled 50 µm measures 20 mm on the screen. A cell measures 34 mm across. Find the real width of the cell.
Step 1: what is 1 mm on screen worth?50 ÷ 20 = 2.5 µm per mmStep 2: scale up the measurement34 × 2.5 = 85The cell is 85 µm wideMeasure the scale bar yourself. Never assume the image is printed at the size the original was.
What technology will not do for you
Bad data in, bad results out. A spreadsheet cannot tell that one reading was taken with the probe out of the water.
It has no sense of significant figures. A mean of 25.83333 from readings taken to one decimal place is false precision — round it to match your raw data.
It will average the wrong cells without a word of warning. Check one result by hand against your calculator.
Spreadsheets reformat things. Readings that look like dates get turned into dates, and units typed into number cells stop the maths working. Keep units in the column heading.
A graph is not a conclusion. The computer draws the trend; you still have to explain it with biology.
💡 Exam tip
Categoric independent variable gives a bar chart; continuous gives a scatter graph with a line of best fit.
Independent variable on x, dependent on y, units in every axis label. Free marks, often thrown away.
Quote a mean with a measure of spread — range or standard deviation.
Say what your error bars represent in the figure caption.
Use the overlap sentence when comparing two means: overlapping bars mean the difference may be chance.
Round processed values to the precision of your raw data, not to whatever the screen shows.
In the IA, keep raw data in one table and processed data in another, with a sample calculation shown.
⚠ Common mix-up
Joining the points dot-to-dot and calling it a line of best fit.
Bar charts for continuous data, which hides the shape of the relationship.
Reporting a mean with no spread, so the reader cannot tell whether the repeats agreed.
Assuming non-overlapping error bars prove causation. They suggest the difference is real; the explanation is still down to you.
Copying every decimal place the spreadsheet gives you. That is false precision.
Extrapolating a straight line off the end of the data, where the biology has almost certainly changed.
Typing over raw readings with processed values, so the original data is gone for good.
Up next: Maths Skills for Biology — the calculations, averages and statistics that sit behind everything you have just processed.
Want this explained one-to-one?
Book a free session with an experienced IB Biology tutor and get your trickiest topics made simple.
