As a Contemporary Study, this study is one of three (the others are Steyvers & Hemmer and Sebastián & Hernández-Gil). This means the Examiner doesn't know which Contemporary Study you will have covered - so there will never be a question specifically on Schmolck et al. You won't get asked what "MTL+" means or how the controls scored in the category sorting task. However, you will get general questions about the Aim, Procedure, Results and Conclusions (APRC) of the Contemporary Study and you may be asked to evaluate it in terms of Generalisability, Reliability, Applications, Validity or (possibly) Ethics (GRAVE).

SCHMOLCK ET AL. (2002)
THE CONTEMPORARY STUDY: BRAIN DAMAGE IN PATIENT H.M.

This study was carried out by Dr Heike Schmolck on a group of patients who had all experienced brain damage and loss of memory.

The most famous patient in the study was “H.M.” – real name Henry Molaison (1926-2002) – who had brain surgery for his epilepsy in 1953 which involved the removal of the hippocampus and suffered amnesia as a result. H.M. would recall information so long as it was in his short term memory, but then forgot it within seconds and could not create new episodic memories. However, he still remembered some things from before his brain damage.

HM has been called the most important patient in the history of brain science. Schmolck wanted to compare H.M. to other patients with similar brain damage to see if a precise link could be made between brain structure and semantic memory.

This study is significant for students in other ways:
  • It shows how scientific research proceeds, because Schmolck used state of the art brain-imaging techniques to identify the parts of the brain damaged in each patient.
  • It focuses on semantic memory, introduced in Tulving's theory of Long Term Memory and the Classic Baddeley (1966b) study.
  • It illustrates features of the Cognitive Approach, since it uses the experimental method to try to isolate and measure semantic LTM
  • It illustrates the power and shortcomings of the natural (or quasi-) experimental method, because the patients’ brain damage was a naturally-varying IV outside Dr Schmolck’s control
  • It shows the growing importance of neuroscience – the link between Cognitive and Biological Approaches

PATIENT H.M. & OTHERS

H.M. underwent brain surgery in 1953 for his life threatening epilepsy. The surgeon removed a part of the brain called the hippocampus. A side-effect was that H.M. was unable to encode new LTM. However, although H.M. lost his episodic memory, he still had procedural memory (such as how to write) and could encode new procedural memories (he learned to play tennis – but he couldn’t remember being taught it).
There were 14 patients in total. 3 (including HM) had brain damage to the hippocampus (part of the medial temporal lobe or MTL) from surgery or other injuries, 3 had brain damage from viral infections (herpes simplex encephalitis) and their brain damage was more widespread – these were called the MTL+ group. 8 of them were Controls who were healthy volunteers with no brain damage. They were matched with the patients in terms of age (70s) and education.

  • Notice Schmolck’s use of the matched pairs design. Each healthy Control was matched against a brain-damaged patient. This is important for any experiment with independent groups, but especially natural experiments where the researcher can’t choose how to assign participants to conditions.
  • Also notice the different types of brain damage. One of the problems with studying “lesions” (damage to parts of the brain) is that patients usually have lesions in several parts of the brain, not just one. The patients with viral brain damage had more extensive lesions than the patients (like H.M.) who had received surgery in just one part of the brain, the hippocampus. However, H.M. had more widespread lesions than the other MTL patients, but not as widespread as the MTL+ patients.

SCHMOLCK ET AL.'S STUDY
APRC

Aim

To find out if Semantic LTM is linked to a particular part of the brain. If so, patients with lesions in that part of the brain should underperform at tests of Semantic LTM. Schmolck focused on damage to the medial temporal lobe (MTL) and the hippocampus and looked in detail at the performance of Patient H.M..
The purpose of the study was twofold: first, to assess the relation between test performance and the extent of damage to lateral temporal cortex and second, to determine whether any aspects of H.M.’s performance were unique among the patients tested - Heike Schmolck
Yes, the MTL/hippocampus is the same brain structure studied by Raine et al. in the Biological Classic Study. You will find out more about the hippocampus when you study the structure of the brain.
IV

The extent of brain injury: (1) 3 patients with damage to Hippocampus/MTL only; (2) 3 patients with damage to MTL and the temporal cortex too (the MTL+ group); (3) a Control group with no brain damage.

Since the IV is naturally-varying and the Controls were matched on age and education, this is a natural experiment with matched pairs design.

In addition, Schmolck used different types of cognitive tests on the patients. The type of cognitive test is a repeated measures design because each participant did every test.

Since H.M. also had more widespread brain damage than the other Hippocampus/MTL patients (brought on perhaps by his earlier epilepsy, perhaps by the less accurate brain surgery in the 1950s) he was also considered separately from the others.

DV

Scores on 9 separate tests of semantic LTM

Sample

6 patients with severe damage to the MTL and 8 Controls with no brain damage. 3 of the patients also had damage to the temporal cortex generally (MTL+).
Procedure

Dr Schmolck created 9 tests for Semantic LTM functions. All were based on a set of 48 drawings, half of animals and half of objects. These pictures were grouped in sixes: 6 land animals, 6 birds, 6 musical instruments, 6 vehicles, etc. Here are some examples:
  • Similar pictures: the participants are shown 6 pictures sharing a theme and asked to point out the one that the researcher names (this is testing for confusion caused by semantic similarity – similar to the Baddeley study)
  • Category fluency: the participants were asked to give as many examples as possible from each theme within a minute
  • Category sorting: the participants were given all 48 pictures and asked to sort them into “living” or “man made”
  • Definitions: the participants were shown a picture and asked to define it by the theme it fitted into

Dr Schmolck also used additional tests that are used with dementia patients. One of these involved 30 pictures that showed either real objects or non-objects: participants had to say whether the object in the picture was real or not.

The participants were tape recorded and their responses transcribed (typed up). 14 'raters' checked each transcript for reliability and also looked for grammar/syntax errors in the way the participants spoke because problems with language also indicates trouble with semantic memory.

Results

Dr Schmolck collected scores for all 9 tests. Here are some examples of her findings:
  • Similar pictures: the Controls got all the answers right as did those with hippocampus damage only (H.M. score 98% for living creatures and 100% for objects); MTL+ patients performed worse: 85% for living creatures and 90% for objects
  • MTL+ Group: These patients did significantly worse in all the tests (p<0.005)
  • Patient H.M.: H.M. did better than the MTL+ patients but slightly worse than the other MTL patients who had damage solely to the hippocampus
  • Overall: Controls scored 99%, MTL patients (excluding HM) scored 100% and MTL+ patients scored 78%
Here's an example of the bar charts Schmolck uses to display her results. It shows the mean scores for correctly identifying living and non-living things from the pictures. CON are the Controls; HF are the MTL patients with damage to the hippocampus only; MTL+ are the 3 patients with widespread lesions; individual scores are black dots and H.M.'s scores are shown as boxes
There was also a positive correlation between the amount of brain damage and the number of mistakes. The MTL+ patients made the most mistakes, followed by HM, then the hippocampus-only patients.

Where the hippocampus-only patients did better than the Controls, Dr Schmolck suggests it is because they were younger (even though Controls were supposed to be matched on this...).

Conclusions

There seems to be a clear link between damage to the temporal cortex generally and the loss of semantic LTM.

Patients with damage specific to the hippocampus suffered loss of episodic memory, but not semantic memory.

This suggests that semantic and episodic LTM are encoded in different parts of the brain, with the hippocampus/MTL dealing with episodic memory and the nearby temporal cortex dealing with semantic memory.

H.M. performed similarly to the Controls but with odd lapses when it came to defining things, where he resembled the MTL+ group more. This suggests H.M. had very specific brain damage that wasn't quite like the others. However, there are problems with generalising from H.M. because of his unusual background. H.M spent most of his young life with debilitating epilepsy and his entire adult life in hospitals, being studied by psychologists.
H.M. had seizures beginning at age 10 (which raises the question whether his language development was fully normal), his schooling was interrupted, and he came from a low socioeconomic background. Any of these factors could be important - Heike Schmolck

EVALUATING SCHMOLCK ET AL. AO3
GRAVE

Generalisability

Schmolck et al. used a small sample – only 3 patients (including HM) with MTL/Hippocampus damage and 3 with wider temporal cortex damage. Samples this small are easily distorted by anomalies – and H.M. seems to have been the anomaly here.

However, Dr Schmolck did single H.M. out as an anomaly because of his wider brain damage. The researchers analysed H.M.'s results in more detail.

These brain lesions and memory problems are relatively rare. H.M. suffered from serious epilepsy. The MTL+ patients all suffered from herpes. This might make them unrepresentative of the wider population.

Reliability

This is a good example of a reliable study because it has standardised procedures that could be replicated by other researchers. MRI scans are becoming common. Dr Schmolck also used 14 raters to check the participants’ scores and their agreement gives this study inter-rater reliability.

Nonetheless, the participants themselves are hard to replicate. HM died in 2005, so no more studies can be carried out on him.

Application

The main application of this study has been for other Cognitive Psychologists, who have built on Dr Schmolck’s research, and earlier studies involving H.M., to understand the brain’s role in memory. This is leading to the development of neurocognitive psychology – a mixing of the Cognitive and Biological approaches.

The study also helps us understand the risks of brain surgery and the side-effects of brain damage, which would enable doctors and patients to weigh up the risks of surgical procedures (HM might not have agreed to his surgery in 1953 if the consequences had been understood).

In the future, this sort of research may even lead to a cure for patients (like Clive Wearing) with this sort of memory loss if brain lesions can ever be repaired.

Because Dr Schmolck used tests that were originally for dementia patients, this might be useful for your Cognitive Key Question.

Validity

The use of healthy Controls and the matched pairs design increases the validity of this study. It means we can be reasonably sure that the different scores on the tests were caused by the brain lesions in different parts of the brain, not by age or intelligence. This is backed up by the MRI scans which showed the temporal lobe area activating when patients had to make semantic judgements.

However, the ecological validity of this study is not good. Naming and categorising drawings on cards is more like a game or a puzzle than the sort of memory you need in real life. It is an artificial test. Other studies (eg Teng & Squire, 1999) asked these patients to recall the neighbourhoods they grew up in and found they performed well at that.

Ethics

There are usually no significant ethical issues with Cognitive studies into memory. However, this study involved patients who could not give valid consent, because they would not be able to remember having the study explained to them. This means the researchers could only obtain presumptive consent from the patients' doctors and carers. HM was studied all his adult life and could never consent to any of it, making him a human guinea pig. Increasingly, ethicists regard it as unacceptable to treat the lack of refusal as tacit consent.

However, the benefits to our scientific understanding of brain functioning from studying patients like H.M. have been so enormous, it may outweigh the lack of consent from these few, unusual patients. This would be a case of a cost-benefit calculation, with one principle of ethics (social responsibility for the common good) counting agaist another (failure to respect dignity and autonomy).

EXEMPLAR ESSAY
An 8-mark essay on the Contemporary Study

Evaluate the contemporary study from cognitive psychology. (8 marks)
  • A 8-mark “evaluate” question awards 4 marks for AO1 (Describe) and 4 marks for AO3 (Evaluate). Remember to include a conclusion in order to get the top band (7-8 marks).
Schmolck et al. had a very reliable experiment. A lot of the tests had been used before with dementia patients and were easy to replicate. She also used 14 raters who studied transcripts of the answers. This gives the study inter-rater reliability.

Dr Schmolck improved the validity of her study by using a control group of healthy adults the same age and educational background as the patients. This is Matched Pairs design. However, the MTL patients did slightly better than the Control group. Dr Schmolck thought this was because they were younger and better educated so the matching wasn’t perfect.

The study is valid because it is backed up by MRI brain scans which show the temporal cortex activating to do semantic tasks. This explains why the MRI+ group (with damage to the wider temporal cortex) scored lower at these tasks.

However, the tests lacked ecological validity because they were artificial. Instead of naming pictures of animals and household objects, the patients could have described their childhood memories like in the study by Teng & Squire (1999).

In conclusion, Schmolck et al. studied patients suffering from an unusual condition. The patients were so unusual that it might be difficult to generalise the results. Now that HM is dead, no more research can be done on him. However, the insights into memory provided by his condition may one day help to cure memory loss and dementia which might make his terrible condition worthwhile.
Notice that for a 8-mark answer you don’t have to include everything Schmolck et al did. I haven’t mentioned the ethical issues or the fact that the sample was too small. But I have tried to make the two halves – Description and Evaluation – evenly balanced.