Laboratories
Statistical Methods in Laboratories
A programme in the statistical tools an analyst needs to evaluate and interpret laboratory data: descriptive statistics and the normal distribution, confidence intervals, significance tests, outlier detection, analysis of variance, and linear regression for calibration curves, applied in spreadsheets.
Overview
About the programme
Duration
4 days
Training hours
12 hours
Delivery
Training hall · Virtual · At your premises
Language
Arabic or English
Objectives
Training outcomes
- Calculate measures of central tendency and dispersion and present data graphically.
- Apply the normal distribution and calculate confidence intervals for the mean.
- Choose and apply the right significance test (t and F) to compare results, methods and analysts.
- Detect outliers with the Grubbs and Dixon tests and deal with them.
- Apply one-way analysis of variance to estimate the components of precision.
- Build a calibration curve by linear regression, assess it with residuals, and calculate unknown concentrations and the detection limit from it.
Content
What the programme covers
Day 1: descriptive statistics and distributions
- Types of data and error, significant figures and rounding.
- Mean, median, standard deviation and relative standard deviation, histograms and box plots.
- The normal distribution and the t distribution, and confidence intervals for the mean.
Day 2: significance tests and outliers
- The null hypothesis, significance level, and type I and type II errors.
- t-tests to compare a mean with a reference value, to compare two means and for paired data, and the F-test to compare variances.
- The Grubbs and Dixon tests for outliers, and the rules for rejecting them and documenting the rejection.
Day 3: analysis of variance and precision
- One-way analysis of variance and interpreting the ANOVA table.
- Estimating repeatability, intermediate precision and reproducibility from designed experiments, and an introduction to ISO 5725.
Day 4: regression and calibration
- Least-squares linear regression, the correlation coefficient and its limits, and residual analysis.
- Calculating an unknown concentration and its uncertainty, and the limits of detection and quantification from the calibration curve.
- Integrated exercise on laboratory data in spreadsheets, followed by the assessment.
Who it is for
Target audience
Analysts, chemists and quality officers in testing and calibration laboratories, and those who prepare method validation and quality control reports.
Method
Delivery method
Short lectures and exercises that every participant solves in Excel on their own computer, using real laboratory data. Data from the organisation’s laboratory can be used if shared before the programme. Delivered in person or online, in Arabic or English.
Questions
FAQ
Does the programme need specialised statistical software?
No. All exercises are solved in Excel with its built-in functions and the Analysis ToolPak, so they can be applied after the programme without additional licences.
What level of mathematics is required?
Basic mathematics and working knowledge of spreadsheets. The programme starts from the basics and concentrates on choosing the right tool and interpreting its result rather than deriving equations.
Is there a certificate?
Every participant who attends the programme and passes the assessment receives a certificate of completion from Technology Experts Training Center, licensed by the Technical and Vocational Training Corporation.
A programme tailored to your organisation
The duration and topics of the programme are adjusted to the organisation's needs, and it is delivered at its site, in a training hall arranged by the Center, or virtually.