The delayed matching-to-sample (DMS) test, also referred to as delayed match-to-sample (DMTS), is a fundamental procedure used to assess cognitive function, particularly visual recognition memory and short-term visual matching ability. This test is employed across various fields, including human neuropsychology, animal cognition research, and behavioural science, to probe working memory and perceptual discrimination under conditions of delay. The procedure involves the presentation of a sample stimulus, which the participant must later identify from a set of choice stimuli after a variable delay interval. The test’s adaptability allows it to be used with non-verbalisable patterns, making it suitable for cross-species comparisons and for assessing cognitive function in individuals who may have language impairments.
The Core Procedure and Task Format
The fundamental structure of the delayed matching-to-sample test is consistent across its applications. The process begins with the presentation of a sample stimulus. In human assessments, such as those administered via the CANTAB® cognitive assessment platform, this sample is often a complex, abstract visual pattern that is difficult to verbalise. The participant is shown this sample for a brief period. After the sample is removed, there is a delay interval. Following this delay, a set of choice stimuli is presented. The participant’s task is to select the choice stimulus that exactly matches the original sample. The test measures the accuracy of this selection, which is taken as an indicator of visual recognition memory and the integrity of short-term memory storage.
The difficulty of the task can be systematically manipulated by varying the length of the delay interval. In zero-delay conditions, the choice stimuli are presented immediately after the sample is removed, testing basic visual matching and perception. As the delay interval increases—commonly to 4 or 12 seconds in clinical assessments—the cognitive load on working memory increases, making the task more challenging. This parametric manipulation of delay allows researchers and clinicians to assess the decay of memory over time and to identify deficits that may only become apparent under higher cognitive load.
In animal research, the procedure is often implemented using operant conditioning chambers. For example, in studies with rats, the sample phase involves the presentation of a specific lever (left or right). Pressing the sample lever results in a reward (e.g., a sucrose pellet) and initiates a delay period. During the subsequent choice phase, both levers are presented, and the subject must press the lever that was presented during the sample phase to receive another reward. The delay duration is randomised to prevent anticipation strategies, and performance is measured by the accuracy of lever selection across varying delays.
Applications and Outcome Measures
The delayed matching-to-sample test is a versatile tool with applications in multiple domains. In clinical neuropsychology, it is used to assess cognitive function in various populations. The CANTAB® DMS test, for instance, is recommended for evaluating cognitive function in individuals with conditions that may affect visual memory and executive function, such as neurodegenerative diseases, psychiatric disorders, and brain injuries. The primary outcome measures include the percentage of correct responses, which provides a quantitative measure of visual recognition memory, and the ability to maintain performance as the delay interval increases, which reflects working memory capacity and resistance to interference.
In animal cognition research, the DMS task is a key paradigm for studying memory processes across species, including pigeons, primates, and rodents. It allows for the investigation of the neural substrates of working memory. For example, a study using rats in an operant DMS task examined the recruitment of neural circuits under low and high working memory load conditions. The research found that high memory load (a 24-second delay) up-regulated c-Fos mRNA expression in the prelimbic cortex (PrL), a region of the prefrontal cortex, and that this activity negatively correlated with choice accuracy. This highlights the test’s utility in linking behavioural performance to underlying neural mechanisms.
The test is also used to study cognitive development and the understanding of relational concepts in non-human animals. Research has explored how animals like gorillas and orangutans understand first- and second-order relations using matching-to-sample procedures. Furthermore, the delayed matching-to-sample paradigm is considered a tool for assessing not only memory but also other cognitive processes, such as attention and discrimination learning, in species like pigeons.
Variations and Methodological Considerations
Several variations of the basic DMS procedure exist, tailored to specific research questions or populations. The "zero-delay" matching-to-sample task, where choice stimuli appear immediately after the sample is removed, serves as a control condition to isolate perceptual matching from memory-based processes. The "simultaneous matching-to-sample" task, where the sample and choice stimuli are presented at the same time, assesses basic discrimination abilities without a memory component.
Methodologically, the test is valued for its non-verbal nature, which reduces the influence of language ability on performance. This makes it particularly useful for assessing individuals with aphasia, developmental delays, or in cross-cultural and cross-species research. The use of abstract, non-verbalisable patterns in human assessments minimises the use of verbal strategies for remembering the sample, thereby isolating visual recognition memory.
The historical roots of the procedure date back to the late 18th century, with early uses in comparative psychology. The basic match-to-sample procedure was used as early as 1799, and the delayed variant was developed to introduce a memory component. Over time, the paradigm has been refined and standardised, especially for clinical and commercial use in platforms like CANTAB®, which provides a computerised, automated administration of the test with precise control over stimuli and timing.
Interpretation of Performance
Interpreting results from a DMS test requires consideration of the overall performance pattern. A general decline in accuracy as the delay interval increases is expected in healthy functioning, reflecting the natural decay of short-term visual memory. However, a steeper decline or poor performance even at zero delay may indicate underlying cognitive deficits. In clinical settings, performance is compared to normative data for the individual’s age and demographic group to identify impairments that fall outside the normal range.
In animal studies, performance is often analysed in relation to specific brain regions or molecular mechanisms. The correlation between neural activity (e.g., c-Fos expression) and behavioural accuracy helps to elucidate the neural basis of working memory. For instance, the negative correlation between PrL activity and choice accuracy under high load suggests that over-activation in this region may be detrimental to performance under demanding conditions, possibly due to inefficient processing or neural noise.
Conclusion
The delayed matching-to-sample test is a robust and widely used paradigm for assessing visual recognition memory and working memory across species and in various clinical contexts. Its strength lies in its simplicity, adaptability, and non-verbal nature, which allow for direct comparisons between different populations and species. By varying the delay interval, the test can probe different levels of cognitive load, providing insights into the capacity and stability of short-term memory. Whether administered to humans via computerised platforms or to animals in operant conditioning chambers, the DMS task remains a cornerstone in the cognitive and behavioural sciences for investigating the fundamental processes of memory and perception.