Robert W Brown

RMIT University, PO Box 2476V, Melbourne 2001, Australia



Whilst much attention is paid to the quality and the avenues of presentation of educational material, there seems to be little consideration given to human physiological processes of information ingestion and assimilation when planning the educational experience.  Human visual and aural information input is highly compressed in a lossy process.  Subjective perceptions relating to past experiences are stored rather than actual data.  Data assimilation rate is stunningly low in terms of learning, making many classroom endeavours futile.  Information is stored, according to its nature, in localised areas of the brain, and involves the consumption and eventual localised exhaustion of reagents such as calcium, sugar and oxygen. Periodic relief is required for replenishment.  Presenting the information in different ways through different channels augments retention and understanding,  Repetition is essential to firmly establish neural links or memory. In this document, a brief review of such factors and how they might be considered in designing learning experiences will be presented.



Humans can read plain English text at a maximum of around 500 words, or around one A4 page of 12 point Times text, per minute [1].  If we consider, for a first cut, that plain English text consists of 27 characters (26 plus space), the information content is 4.76 Bits/character.  Redundancy in plain English reduces the information content to about 1 bit per character. [2].

A quick check of the author’s lecture notes indicates an average of around six characters per word.  Using this number as the average number of characters per word, give around 3000 bits per minute, or 50 bits per second as the peak rate that humans can read.  The average human will not read this fast.  In addition, such reading does not entail full assimilation and processing of the information.  Raisbeck [1] estimates that humans can absorb only around 0.2 Bits per second over periods of many hours.  The notes handed out by this author in an undergraduate engineering course, typically contain around 20,000 characters per lecture hour.  This corresponds to an information transmission rate of around 5.4 Bits per second.  This is 27 times the long-term bit-processing rate of humans estimated above.


Input Channels: Roughing Out the Limits

If we consider that the aural bandwidth of human hearing is around 3000 Hz and the signal to noise ratio is typically around 20dB, the channel bandwidth is around 20,000 Bits per second.  Estimating visual data processing rate is more difficult. Humans are able to perceive and react to very fast moving objects such as baseballs or unfolding road views.  However, this is largely autonomous and does not appear to involve conscious processes to a large degree.  The same can be said of a professional typist or stenographer.  Frequently, they do not appear to consciously know where particular keys are but “their fingers know.”  Hence if asked, they need to rest their fingers on the key to rapidly describe where a particular key is located.  Fairly clearly in this case, the visual signal is routed through the brain directly to areas that control the fingers.  Processing a normal image so that it can be described later takes much more time.  If we assume without too much justification, that typically a normal person would need perhaps, ten seconds to “memorise” any image and each image is composed of around 1 MBit of information, then the corresponding visual channel bandwidth is around 100,000 Bits per second.

If the brain were able to store all this information continuously for a lifetime of 70 years of 16-hour days, around 1.5E14 Bits of memory would be required.  This exceeds the typical available brain neuron count by a factor of between 10,000 and 100,000.  (Many neurons are committed to other senses and to controlling body function and hence not all are available for vision.)  If we assume that each neuron, in its complex web of synapses, corresponds to one bit of storage capacity, it is clear that inadequate storage is available and the brain does not everyday store such data and only fleeting correlations of what is seen and heard can be stored.

If one uses a reduction factor of 100,000 on the above numbers to reduce the effective data input rate to what might be processed and stored, the aural processing capacity is reduced to 0.2 Bits per second and the visual processing capacity to 1 Bit per second.  This is of the same magnitudes as has been determined experimentally [1].

This correspondence tends to support the contention that we may rely on a comprehension and mastery rate of no more than 1 Bit per second and probably as low as 0.2 Bit per second in long term learning processes.  This means, that the average student will require between 20,000 and 50,000 seconds to learn the material in the notes that are typically given for one hour of lecture by the author of this paper.  This corresponds to 5.6 to 28 hours.  Since the students typically have 10 hours per week of lectures, somewhere between 56 hours to 280 hours per week of extra study may be required, an impossible situation.  Thus poor learning outcomes that often puzzle despairing academics may be simply a consequence of trying to pump too much information down a band-limited channel in too short a time.  It should be noted that we are talking about assimilation of data into meaningful and accessible memories, not the ability to observe detail.  Observation is at a much higher rate but is not usually committed to memory or assimilated.


Information Content and Storage

Typical lecture notes require around 2000kbits of computer storage memory per hour of lecture.  An unknown proportion of this is formatting and other data needed by the word processing package but at least this figure provides an upper boundary for total information content.  The notes contain a mix of text, numbers and graphics.  Redundancy in English text will further reduce the actual information content.  The information content represented by the characters at one bit per character, is around 20kbits per hour of lecture.  Depending on the relative weighting of text and graphics in a written article, the high information content of graphics can be expected to raise the total information content of an hours lecture, above the 20kbits corresponding to text alone.

Thus the total information content of lecture notes for an hour of lecture will be above 20kbits but less than 2000kbits.  However, experience tells us that “a picture is worth a thousand words” and that such graphics, even simple line drawings, greatly enhance the teaching and learning process.  “Information” is conveyed much more readily with the deployment of embedded illustrations and the material is not so “dry.”  The habit of the brain of interpreting information in the light of past images and experience implies that the raw information is not stored, but that to a large degree, interpretations of the information and images are remembered.  Different people have different interpretations and so they perceive things differently.  The police have long regarded as suspect, eyewitness accounts because of this difference in perceptions.

The key factor is that the information is compressed and stored in an efficient but lossy way.  The information is at least partially interpreted as correlates of previously stored images and knowledge rather as sprites are used in MPEG7.  Thus the real data storage rate is much less than would be required to store the raw data.  This is a very effective and efficient compromise but the price paid is the opening of door for errors and misinterpretations.


Learning and Storage Processes

The physical actions of individual memory cells in forming memory or data storage, is fairly well understood [3].  Short-term memory may be thought of as new and fragile linkages between neurons.  Consolidation of these to form long-term memory is essentially dependent on the hippocampus and medial temporal lobe [4].  The memories are stored at various locations in the brain, including the cerebral cortex.  The hippocampus supplies critical hormones whose abundance declines with age.  Thus older people often have perfect recall of events long passed, recorded at younger ages, but short term memory and the formation of new long term memories is inhibited by the shortage of these critical hormones.

Physical processes must occur for memories to be stored.  Calcium ions are transferred between neurons in forming new memory.  Significant increases in the rate of consumption of sugar and oxygen occurs in the region of the neural activity.  In short, chemicals are consumed in the operation of the brain, especially in forming memories.  The statement “my brain hurts” jestingly made by students after a full-on and mentally demanding lecture or study exercise, in a way, is therefore likely to have a physical basis as is the complaints of “mental exhaustion.”  The depletion of vital chemicals in the brain by intense activity needs to be taken into account when structuring learning exercises.  Giving students a regular break during a course is not then an act of charity catering to an abstract psychological need, but a sensible requirement based on physiological reality.



Modern electronic equipment such as PET (Positron Emission Tomography) scanners and NMR (Nuclear Magnetic Resonance) imaging devices, have allowed unprecedented unravelling of the actions of the brain [3].  Static images are stored in one area, moving images in another; verbs are processed in a different area to that for the processing of nouns.  Information that is written is stored in different ways to information that is spoken.  In short, the storage and recall mechanisms are very diverse and complicated. 

There are two important outcomes of this.  The first is that the same information presented through different channels and deployed by the student in different ways such as writing, speaking and conducting of exercises, will be stored in different parts of the brain in a way that reinforces and consolidates the learning experience.  The second is that repetitive presentation and manipulation of the information is vital to the memory process.  Rote learning, often disparaged, thus can be a critically effective part of the processes needed to reinforce linkages between neurons, commonly referred to as neural networks.  These networks are memory and are retained and strengthened only through repetitive use.

Toru Kumon [5] created a method based on rote learning for his son in Osaka in 1954.  Critics referred to it as “drill and kill” but tests showed that all the mathematics students who used it benefited from its use, irrespective of their native abilities.  It can be argued that such memorisation only addresses the first of the six levels of Bloom’s Taxonomy [6] – that of knowledge.  However, practice is not restricted to repetition of facts, but can be applied to all six stages – knowledge, comprehension, application, analysis, synthesis and evaluation.

In any case, learning and memory are inseparable according to Lemonick [3].  He succinctly states:  “We think of learning and memory as separate functions; in fact they are not.  Both are processes by which we acquire and store new data that makes them retrievable later on.”  Initially, new information is usually held in short term memory and it is essential that nascent learning be reinforced within a short time frame, less than a day for example, for the learned information to be retained.  Thus laboratories, project work or, less effectively, recapitulation of the information previously covered, is an essential part of obtaining stimulation and locking in the memory or learning experience.  Timely positive feedback is also important to enhance retention and reinforcement of the new memories.


Perceptions and Interpretations

Information received by the eye is pre-processed in the eye before transmission to the visual cortex.  Ongoing processing extracts, for example, edges, movement and colour.  The eye has 100-200 million photoreceptors but the optic nerve connecting the eye to the primary visual cortex, has only a million nerve fibres [7].  Therefore, data compression of around 100:1 is done in the complex interconnections within the eye.  The compression is not uniform across the retina, but tends to be high in the peripheral vision and quite low in the fovea.

After processing in the visual cortex, the information is coded and stored in different parts of the brain. [3], [8]  In short, the visual information is not simply stored as in a computer memory, but is highly processed and compressed before storage.  Further, if we accept that the brain tends to store interpretations and impressions of incoming information, then it is not surprising that, in terms of a lecture presentation for example, students register widely different interpretations of what was presented.  Therefore it is also not surprising that lecturers are sometimes astonished to find that students have totally “misinterpreted” what he or she thought was being conveyed to the students.

What seems perfectly clear to the academic may not be clear at all to the students, because their interpretations of the information are founded on an entirely different catalog of experiences.  Further, this catalog is much more limited by their lack of experience.  This provides fewer prior memories on which to attach new information and concepts and thus learning is initially restricted in breadth and capability.  It is a bootstrapping situation.  New information and images are slowly built up which enhances the subsequent learning process.  An example of differences in interpretation is given in Figure 1.

A further danger of the brain’s memory and information processing mechanisms is clearly represented by false memories and the habit of the brain of “filling in the blanks.”  Loftus [9] clearly describes how the brain can be easily made to accept false recollections of information that was never presented.  Whilst having disastrous effects on some families in criminal actions over events that never actually occurred, the same process can cause the educational process to be undermined.  Critical analysis, problem solving and teamwork can be crippled by perceptions of “facts”, strongly believed, but false.  Such strongly held beliefs can frustrate the efforts of academics in promoting the development of skills in students.

“Filling in the blanks” can occur when information is lacking or the information presented is ambiguous or incomplete.  As described by Lemonick, [3] the brain, particularly through vision, will often see things that are not there or will see different things in the same image. An example is given in Figure 2.



Spiegal in a quote by Concar [10] declares:  “All perception is a combination of some raw sensory input and some internal mental image or concept. In hypnosis we’re setting up a competition between the two … we’re saying use the internal image to change what you see.”

Hypnotists use this to advantage to create illusions.  Conceivably, the same process could be used in education but the triggers would need to be very carefully conceived and administered.  In practice we take little account of such processes.  We do note its importance in society by saying to people aspiring to higher office, that “first impressions are everything” and “image is very important – dress to impress.”  Such triggers conjure up a positive impression of competence and energy.

The upshot of such variations in images and perceptions, is that academics must not assume that what they are presenting in their view is the same as the one perceived by the students.  It is important to ensure that the information is presented in a number of ways and accents until there is a convergence of student’s interpretations and the intended outcome in presenting the data.



Certain images and events are imprinted in most people’s minds.  The assassination of John F Kennedy (1963), the space shuttle Challenger disaster (1986) and the death of Princess Diana (1998), are events that people not only recall very clearly, but also correlate with recall of exactly what they were doing at the time of the event.  The operation of memory is clearly excellent in these cases and points to the impact of neuro-biological factors in the formation of memory.  Emotions trigger the release of chemicals and hormones that greatly facilitate memory effectiveness by their action on neuro-transmitters and on other factors such as blood flow and oxygenation.

Adrenaline is one such chemical (a steroid) that enhances the learning experience, especially if you are being chased by a grizzly bear at the time.  You will not forget it.  Similarly, in a classroom, the efficiency of learning is strongly affected by the level of engagement of the student.  In short, are they interested and involved?  Students are unlikely to learn much from a lecture that, to them, is boring or uninteresting.  The chemical processes vital to learning will be idling.  It is essential to rev up interest if the educational process is to be effective.  In older days, this could be done through imposed discipline and fear but other methods must now be used.  Physical activity or participative learning is one common very effective method of obtaining such engagement.  Such processes also engage the vital complementary learning processes through the deployment of different parts of the brain through the use of different channels and actions of the student.



Human beings are not simply information storage systems resembling tape recorders.  Information digestion rate is severely limited to 1 bit per second or less over the long term.  This should be taking into account when deciding on the content of learning exercises.  Humans use complex electro-chemical processes and extreme, lossy data compression in processing and storing information.  The information stored is interpretations of the data observed leading to differences in perception.  The deployment of physical reagents in the brain and parallel storage strategies, corresponding to interest, engagement and use of aural, visual, tactile and participative or experiential learning, bring into play the brain’s parallel processing and storage capabilities and are important considerations in the learning process.

The operation of the brain involves physical processes involving complex electrochemistry and the consumption of chemicals such as oxygen, calcium and sugars.  Thus the brain needs a regular respite at least to refresh its reagents and probably for much more complex purposes.  This critical need should be factored into the teaching and learning process, with adequate times allowed for breaks and refreshment.  The propensity of the brain to “fill in the gaps” and to interpret information in the light of past experience must be reckoned with in obtaining acceptable interpretations and perceptions of the information being presented.  Memories constructed of neural networks that owe their existence to timely reinforcement by repeated experiences, point out the critical necessity for repetition of the same information within a short time frame, and the use of different variations on the same theme.



[1] Raisbeck, Gordon, Information Theory, Massachusetts Institute of Technology, 1963.

[2] Shannon, C. E., Weaver, W., A Mathematical Theory of Communication, University of Illinios Press, Urbana, Ill, 1949.

[3] Lemonick, Michael D., Glimpses of the Mind, Time, July 31, 1995, pp 52-60.

[4] Lemonick, Michael D., Smart Genes, Time, September 27 1999, pp 58-62.

[5] Time Magazine, August 7 1995, p 23.

[6] Bloom, B. S., (Ed.) Taxonomy of Educational Objectives. Handbook I: Cognitive Domain, New York, McKay.

[7] Weblin, Frank, Jacobs Adam, & Teeters, Jeff, The Computational Eye, IEEE Spectrum, May 1996, pp 30-37.

[8] Braham, Robert, Toward an Artificial Eye, IEEE Spectrum, May 1996, pp 20-29.

[9] Loftus, Elizabeth F., Creating False Memories, Scientific American, September 1997, 277, Number 3, pp 50-55.

[10] Concar, David, You are feeling very, very sleepy, New Scientist, 159, No. 2141, 4 July 1998, pp 26-31.