The scientific process for chemistry
On this page
Before conducting any experimental work, establishing what you are trying to achieve is essential.
Having a clear experimental aim makes it easier to assess what is important for success.
In chemistry, experiments are primarily investigations or syntheses.
- In an investigation, planning to manage your variables is key.
- In syntheses, focus on confirming product purity and obtaining a high yield.
In A level, it is unlikely you will be directly tested on defining the types of variables, but it is important they are considered in any experimental plans.
You adjust the independent variable to discover its effect on a result. It is the thing you change and is typically found on the x axis of a results graph.
When discussing your independent variable, state how it will be measured and to what degree of accuracy.
Where the experimental aim is to show a trend, discuss the range of values required for your independent variable, and the interval between them.
Where the experimental aim is to obtain an absolute value, consider the number of repeats required to ensure accuracy.
The dependent variable is the thing you observe or measure as an experimental outcome. It changes as a result of the independent variable and is typically found on the y axis of a results graph.
When discussing your dependent variable, state how it will be measured and to what resolution.
Control variables are not part of the relationship under investigation but could impact the dependent variable. Measures are taken to ensure that control variables are kept constant as much as possible.
Where control variables cannot be easily controlled, their values should be measured, recorded, and considered in the analysis of results.
Potential sources of error should be considered in the planning stage.
It is most important to consider any factors which could impact the accuracy or precision of the data collected for the independent or dependent variable.
In chemistry, error prevention methods include:
- Selecting appropriate equipment for the precision requirements. This includes the choice of glassware and decimal point accuracy on balances.
- Establishing when to rinse equipment, what to rinse it with, and whether to transfer washings. Consider the impact on concentrations and moles transferred against your experimental aims.
- Identifying how to take measurements, including any systematic errors that could impact the accuracy of the reading. Consider the positioning of thermometers, the need to stir solutions, factors aiding the visibility of an endpoint, and the viewing angle for the level of a liquid.
Precision and accuracy are both linked to reliability.
Precision relates to how consistently the same result can be obtained. It relates to the standard deviation of repeat measurements. The resolution (smallest discernible increment) of equipment used will impact precision.
Accuracy relates to how close the mean measured value is to the true value. Where there is an offset, this is called bias. Bias is usually linked to a consistent or systematic error.
Structure your results table before beginning practical work.
Include space for all the raw data recorded during the experiment, as well as for relevant calculated differences.
Consider the need for repeat readings when structuring your table.
Often the data required for calculations is the difference between two experimental readings.
Include space to record the raw data, as well as any calculated values.
Adding additional columns or rows into tables for repeat measurements rather than drawing multiple tables:
- Saves time
- Allows the mean to be calculated from a single table
- Makes it easier to see anomalous data
Label each column in a table with the quantity being measured and its unit in brackets.
Column headers could include:
- Time
- Temperature
- Concentration of
All chemistry practicals require a risk assessment.
It is important to consider physical and chemical risks and plan to mitigate them through the appropriate use of PPE.
Physical risks include:
- being burnt by heat generated in the experiment, or by a hot plate/open flame
- breaking glassware and being cut
- dropping heavy items
- building-up of pressure.
Chemical risks are specific to the materials used in an experiment and must consider both the amount of material being used and its concentration.
In a school setting CLEAPSS hazard cards are a good source of appropriate control measures to apply.
In a test setting, you will be expected to recall good lab practice and to identify key additional safety requirements, such as a fume hood being required alongside the generation of a toxic gas.
Analysis of experimental results provides the evidence to support a conclusion.
When analysing results, clearly show the link between what the data show and how this can be interpreted.
The best method of analysis varies according to the experiment.
The most common analytical tools used in chemistry include:
- Scatter graphs
- Tick box (if … then statements)
- Inputting values into equations
A scatter graph is used when a numerical relationship between the dependent and independent variables is being analysed.
One variable becomes the x axis, while a second becomes the y axis. A line of best fit shows the relationship between the variables.
A third variable can be included on the same graph through use of a different marker type or trendline colour.
A good line of best fit should:
- be well balanced. It should have roughly the same number of markers above it as below it.
- map the general trend of the markers rather than connecting them.
- ignore obvious outliers.
- be drawn in a single continuous motion using a sharp pencil.
Do not assume a line of best fit will be straight, but if a linear trend is shown then the line of best fit should be drawn using a ruler.
Some experiments require separate lines of best fit to be drawn for different stages of the investigation.
The extrapolation of both lines of best fit from before and after the second reactant was added is used to obtain the value needed for analysis in experiments using a bomb calorimeter.
The gradient represents the rate at which the variable changes in relation to the variable.
In a linear graph, the gradient is constant for all values of
Where a graph is curved, the gradient changes with the value of . Absolute values for the gradient of a curved graph are only valid for specific values of and can be found by drawing a tangent.
To find the gradient at a specific point on a curve, draw a tangent.
A tangent is a straight line which only just touches the curve at the value of interest.
Two coordinates on the tangent should be selected to calculate the gradient.
It is easier to read coordinates that sit directly on gridlines.
The intercept of a graph is the point at which the line crosses the axis. Graphs can have axis intercepts and axis intercepts.
The axis intercept is the value where .
The axis intercept is the value where .
In a linear equation, , the axis intercept is and the axis intercept is .
Displaying experimental results in a scatter graph and drawing a line of best fit can be used to link data to a known equation.
The linear form of the Arrhenius equation is a good example of how the gradient or intercept on a graph can be used to solve an equation.
When is plotted against and , the gradient is and the y intercept is .
‘X … therefore … Y’ logic is useful in qualitative and spectroscopic analysis.
A well-designed experiment can be analysed by ‘ticking off’ a predefined list of expected observations.
Using ‘X … therefore … Y’ statements, when analysing results in an exam, correlates well to working in the mark scheme.
Experiments linked to amount of substance require algebraic analysis.
Maintaining the link between any values recorded and the sample they relate to is critical.
Annotation of working makes it easier to trace numbers through your calculation and ensures better correlation to method marks in the mark scheme.
When describing the precision of numbers:
- ‘decimal places’ refers to the total number of digits shown after the decimal place.
- ‘significant figures’ refers to the number of significant digits shown after and including the first non-zero digit.
The following rules apply when determining the number of significant figures.
The significance of zeros to the right of the last non-zero digit in an integer (where no decimal place is shown) varies and needs to be explicitly stated. could be accurate to 3, 4, 5, or 6 significant figures.
Significant figures and decimal places are both used to describe the precision of numbers.
Decimal places are used when describing measured values. The reading from a burette or a balance is accurate to a set number of decimal places; this is the equipment’s resolution.
The number of significant figures of measurements is variable depending on the sample size. 54.056 g and 0.002 g are both three decimal places (3 d.p.), but has five significant figures (5 s.f.) and only has one significant figure (1 s.f.).
Significant figures are used when providing calculated values and should reflect the significant figure precision of the input values.
Answers should be given to the number of significant figures stated by the question.
Where the required number of significant figures is not given in the question, an appropriate number should be selected based on the precision of the input data.
The input data with the lowest precision dictates the maximum number of significant figures for an answer.
Where numbers are greater than 10,000 or less than 0.001 they are commonly shown in standard form and to an appropriate number of significant figures.
Any number can be represented in standard form:
where is in the range .
The number of significant figures is the number of digits shown in .
Anomalous results are those which do not fit in with the rest of the experimental data. They can be identified most easily when results are plotted on a graph.
It is common to exclude anomalous results before processing data. This should only be done when the result is thought to come from experimental error; excluding results should not be used purely so data better fits a trend.
All lab equipment has a degree of absolute uncertainty associated with its readings. Unless stated otherwise, this is ±0.5 of the smallest measurement increment.
Absolute uncertainty can be converted to percentage uncertainty by considering the total measured value.
The smaller the measured value, the larger the impact of the absolute uncertainty.
Where a value of interest is obtained from the difference between two measured values, the uncertainties are added together.
Examples of this are weighing by difference and taking start and finish readings on a burette.
Margins of error incorporate all the individual measurement uncertainties as well as other sources of error across an experiment.
Improved accuracy of individual experiments or an increased number of data points can reduce margins of error.
Taking the mean value from repeat experiments is a common way to reduce the margin of error.