Many of the materials utilised in current technologies, especially those used in the aerospace industry, require an unusual combination of properties that cannot be met by conventional metallic, ceramic or polymeric materials alone. Therefore, aerospace structural engineers are constantly searching for alternative materials, such as composites. Composite materials consist of two or more constituents with varying properties, that when used together, form a new material. These can be tailored to have low densities, high strength and stiffness, and have good corrosion resistance. These materials improve structural efficiency, and can reduce the structural mass without compromising structural stiffness and strength. The ultimate objective of the research work, described in this paper, was to conduct an experimental investigation to detect and quantify impact damage in structures made from composite plates. Two different approaches were used to record acceleration response signals resulting from impact excitation. The first used a standard impulse hammer in order to acquire time data and produce spectra for a number of non-damaging impacts. The second approach used a drop-test to perform a number of potentially damaging impacts. The impact energies for the drop tests ranged from 0.37 J to 41.72 J. Following a systematic series of experiments on the induction of impact damage in composite specimens, two different techniques were used to examine the impacted samples. X-ray radiography was used to evaluate the damage area whilst Scanning Electron Microscopy (SEM) was used to inspect the topographies of the specimens at high magnifications. The motivation behind this research work was based on the idea of using statistical or machine-learning methods to identify and categorise impacts (damaging or non-damaging and also types of failure modes) using only structural response data. The discussion here will be focused on outlier analysis on higher-dimensional multivariate data, the features are from binned spectra which leads to an 8-dimensional feature.