Recording Data Honestly

Introduction

An experiment is only as good as the data it produces, and data are only useful if they are recorded clearly and honestly. A number without a unit, a results table without headings or a "tidied up" set of readings can make a careful experiment worthless. This page shows how chemists record, organise and process data so that the results can be trusted by anyone who reads them.

Core explanation

Results tables. A good table has a clear title, column headings that name each quantity and give its unit in the heading — for example "Time (s)" or "Mass (g)" — and the independent variable in the first column. Units go in the heading, so each cell contains only a number. Values should be recorded to the same number of decimal places as the instrument can read; a balance reading to 0.01 g should give values such as 2.40 g, not 2.4 g.

Units. Chemists use the international system of units (SI) and a few related units. Mass is measured in grams (g) or kilograms (kg), volume in cubic centimetres (cm³, which equals millilitres, mL) or cubic decimetres (dm³, which equals litres), time in seconds (s) and temperature in degrees Celsius (°C) or kelvin (K). A number without a unit is meaningless: "5" could be 5 g, 5 kg or 5 minutes.

Repeat readings and the mean. Each measurement is repeated, usually three times. The mean is found by adding the values and dividing by how many there are. Before calculating the mean, check for anomalous results — values that are clearly out of line with the others, perhaps caused by a timing error or a spill. An anomaly should be kept in the table, marked, and left out of the mean with a note explaining why, rather than being erased.

Honesty. Data must never be invented, changed or selectively removed to make results match a prediction. Doing so is scientific misconduct. It misleads other people, can cause real harm when results are used for medicines or safety decisions, and prevents the discovery of genuinely interesting effects. Unexpected results should be recorded exactly as observed and discussed in the evaluation.

Graphs. When both variables are numerical, results are usually shown as a line graph, with the independent variable on the x-axis and the dependent variable on the y-axis, both axes labelled with quantities and units. A line or curve of best fit helps reveal the pattern and makes anomalies easy to spot.

Step-by-step reasoning

To process a set of repeat readings:

1. Record all readings in the table with units in the heading. 2. Look for any reading that differs greatly from the others; mark it as anomalous. 3. Add the remaining readings and divide by their number to find the mean. 4. Round the mean to a sensible number of decimal places, matching the readings. 5. Plot the mean values on a graph and draw a line or curve of best fit.

Visual explanation

Results table for "Effect of temperature on dissolving time":

Temperature (°C) Time 1 (s) Time 2 (s) Time 3 (s) Mean time (s) --- --- --- --- --- 20 120 118 122 120 40 75 78 72 75 60 42 40 71 41 80 25 24 26 25

Anomalous (the stopwatch was started late); excluded from the mean.

Real-world analogy

A results table is like a bank statement. Every entry needs a date, a description and an amount in a clear currency. If someone secretly changes an entry to make the balance look better, the whole statement becomes untrustworthy, even if every other line is correct. Scientific data work the same way.

Real-world example

Pharmaceutical companies must keep complete, original records of every test on a new medicine. Regulators inspect these records before approving the medicine. When companies have been found to hide unfavourable results, medicines have been withdrawn and patients have been put at risk, which shows why honest recording matters far beyond the classroom.

Why?

Why should anomalous results be marked rather than deleted? A value that looks like a mistake might sometimes reveal something real that nobody expected. Keeping it visible, with a note, allows others to judge whether excluding it was reasonable and prevents results from being quietly shaped to fit a prediction.

Common misconception

Some students believe that results which do not match the prediction must be wrong and should be repeated until they "work". Repeating is good practice, but results must be recorded as they are. If the prediction was wrong, the honest conclusion is that the evidence does not support it.

Worked example

Question: Three readings for the volume of gas collected in one minute are 46 cm³, 48 cm³ and 47 cm³. Calculate the mean.

Reasoning: No reading is anomalous. Sum = 46 + 48 + 47 = 141 cm³. There are 3 readings.

Answer: Mean = 141 ÷ 3 = 47 cm³.

Quick check

1. Where should the unit be written in a results table? Answer: In the column heading, for example "Mass (g)", so that each cell contains only a number.

Exam focus

When drawing tables and graphs in exams, marks are awarded for units in headings, consistent decimal places, the independent variable in the first column or on the x-axis, labelled axes with units, sensible scales and a best-fit line. When calculating means, show your working and exclude anomalies with a reason.

Advanced insight

Every measurement has an uncertainty, often taken as half the smallest division of the instrument (for example ±0.5 cm³ on a measuring cylinder marked every 1 cm³). At higher levels you will combine these uncertainties to judge how confident you can be in a calculated result — an essential skill in analytical and physical chemistry.

Summary

Data should be recorded in tables with clear headings, units in the headings and consistent precision. Repeat readings allow a mean to be calculated; anomalous results are marked and excluded from the mean with a reason, not erased. Graphs show patterns. Honest, complete recording is essential because others must be able to trust and check the results.

Practice questions

1. Why is "The mass was 12" an incomplete record? Answer: It has no unit, so it is unclear whether it means 12 g, 12 kg or something else. 2. Readings of 30 s, 31 s, 52 s and 29 s are taken. Identify the anomaly and calculate the mean of the others. Answer: 52 s is anomalous. Mean = (30 + 31 + 29) ÷ 3 = 30 s. 3. Which variable goes on the x-axis of a graph? Answer: The independent variable. 4. Give one reason why changing results to fit a prediction is harmful. Answer: It misleads others, can cause harm when data are used for decisions such as medicine safety, and hides genuinely new findings.