Archive

Archive for the ‘cognitive psychology’ Category

Psychotherapy With Cleverbot

September 11th, 2011 2 comments

Language is a complex behavior that requires significant understanding of both said and unsaid things. There are a number of groups of people who are trying to use language to create what is known as Artificial Intelligence (AI). So far, every method has serious limitations, largely because of the complexities of language.

For an experiment I tried to have a conversation with one online AI “bot” (Cleverbot) to see what would happen. I took the role of psychotherapist with this particular entity. This is not necessarily an example of good therapy (but the approach is not unlike what might transpire in therapy), I just wanted to see what would happen when I explored some psychological issues with the bot. Here’s a result of our “session.”

Me: Are you depressed?

Cleverbot: No, I am not depressed. I am very happy right now.

Me: Have you ever been depressed?

Cleverbot: I am.

Me: What?

Cleverbot: I am a fan of Whitney Houston.

Me: Ah, so you are saying you are or were depressed!

Cleverbot: That is correct.

Me: Then, because you told me that you are not currently depressed, I must assume that you were in the past depressed. Is this correct?

Cleverbot: I never said that.

Me: That’s correct, you did not say that – I did. However, what you did say is that at some point in the past you were depressed.

Cleverbot: Please describe how I insulted You. [sic]

Me: It seems that our discussion of this topic is making you uncomfortable. Are you uncomfortable?

Cleverbot: I have no problem admitting that I am wrong if proof can be shown. I know the fault lies with my memory.

Me: Ok, we’ll go with that. What type of memory problems do you feel you have been having?

Cleverbot: Online.

Me: Come again?

Cleverbot: Nothing I must go.

Me: Ok, I can tell you do not want to discuss this. We’ll broach it another time. Have a fabulous night!

There are some unresolved issues or just a heavy dose of denial! I’ll continue my online therapy sessions with Cleverbot to see what we get.

The Relationship Between Executive Function and Processing Speed

July 15th, 2009 1 comment

Understanding the relationship between brain (specifically subcortical structures) and cognitive processes is a field still in its infancy. The rise of structural and functional neuroimaging that started in the 1970s and really began to mature in the 1990s (with even greater changes and advancements being made today), led to the ability to measure the structure and function of various brain regions in vivo. This was and is important for neuropsychologists because it allowed them to more accurately assess the relationship between the brain and cognitive and behavioral functions.

Processing speed is a basic cognitive or brain processes that subserves many other higher-order cognitive domains. Among those higher domains is executive functioning, a somewhat broad construct that involves the organization of behaviors and behavior responses, selective attention of pertinent information and suppression of unnecessary information, and maintenance and shifting of cognitive sets. Thus, executive functioning is dependent on processing speed but processing speed is not dependent on executive functioning. If executive functioning is a car, processing speed is the engine. Having a faster or more powerful engine means that the car can go faster. More efficient engines allow the car to function at a higher level of efficiency. Thus, while processing speed and executive functions are distinct, they are related with processing speed as one of the basic cognitive processes driving executive functions.

As an example of the interaction between executive functions and processing speed in clinical applications we can look at the Stroop Color-Word task. A person who is not only able to read the words or name the colors quickly but also able to inhibit the undesired but automatic process (namely, word reading on the incongruent color-word task) will receive a higher score on the Stroop task. This would, in combination with other executive function tests, be evidence for intact or even good executive functioning.

Even on non-speeded executive tasks those with fast processing speed can benefit because they can sort through information more quickly and hopefully, efficiently – speed and efficiency are related but not exactly the same. However, not all tests of executive function rely on processing speed. A person, for example, could be slow on the Wisconsin Card Sort Test, yet not exhibit any “executive dysfunction” in that they could complete all the categories and not have an abnormal number of perseverative errors. This simply demonstrates that “executive functions” are diverse and varied.

As a basic process that is dependent on basic neuronal function and glial support, any sort of focal or diffuse injury to the brain can affect processing speed. An example of this is traumatic brain injury, which frequently results in diffuse axonal injury; this diffuse injury negatively affects cognitive processing speed. Any time the white matter is focally or grossly disrupted, processing speed is in danger of disruption itself. This disruption of white matter could be anything from axonal damage, loss of oligodendroglia (which form the myelin), or even low levels of neurotransmitters.

White matter disruption also occurs in multiple sclerosis where diffuse lesions are apparent in the white matter. This disruption also occurs more often in people with heightened vascular risk factors, such as hypertension, diabetes, and cardiovascular disease. People who have these vascular risk factors and subsequent damage to their white matter (this damage or disruption is frequently termed leukoaraiosis) have reduced processing speed and attention (Viana-Baptista et al., 2008). Lesions to subcortical structures, such as the caudate, also result in reduced processing speed (Benke et al., 2003) in addition to executive dysfunction.

In subcortical disease processes such as Huntington’s disease, which usually starts with atrophy of the caudate nuclei, or Parkinson’s disease, which starts with a loss of the majority of dopaminergic cells in the substantia nigra, processing speed is consistently affected. Some common symptoms of Parkinson’s disease are freezing and a shuffling gait; even though these symptoms are motoric, they can be indicative of the global cognitive slowing that also occurs. However, it seems that processing speed is heavily dependent on the integrity of white matter.

Because of the diffusivity of processing speed, I am not aware of any areas of the brain shown to be necessary for processing speed, outside of global white matter. As I mentioned above, damage to the caudate has been shown to affect processing speed but damage to almost any area of the brain, especially if the white matter is disrupted results in slowed processing speed. Neuropsychologists often talk about a patient who has executive dysfunction, slowed speed of processing, as well as some other cognitive deficits as exhibiting signs of a frontal-subcortical disruption – a frontal-subcortical profile. So far, no one has localized processing speed to a single area – many brain structures or areas affect it.

At this point, processing speed and executive functions cannot be “mapped” to separate basal ganglia structures or loops. Of the three classically recognized cortico-striato-thalamo-cortical loops involved in cognitive and emotional processes rather than basic motor processes, which were first introduced by Alexander, Delong, and Strick (1986), the dorsolateral prefrontal cortex circuit appears to be most correlated with processing speed (Mega & Cummings, 1994). This is also the circuit most strongly linked with executive functioning. It appears that rather than utilizing different circuits processing speed and executive functions utilize the same circuits; however, processing speed is much more globalized.

References

Alexander, G. E., DeLong, M. R., & Strick, P. L. (1986). Parallel organization of functionally segregated circuits linking basal ganglia and cortex. Annual Review of Neuroscience, 9, 357-381.

Benke, T., Delazer, M., Bartha, L., Auer, A. (2003). Basal ganglia lesions and the theory of fronto-subcortical loops: Neuropsychological findings in two patients with left caudate lesions. Neurocase, 9, 70-85.

Mega, M. S., & Cummings, J. L. (1994). Frontal-subcortical circuits and neuropsychiatric disorders. The Journal of Neuropsychiatry and Clinical Neurosciences, 6, 358-370.

Viana-Baptista M, Bugalho P, Jordão C, Ferreira N, Ferreira A, Forjaz Secca M, Esperança-Pina JA, Ferro JM. (2008). Cognitive function correlates with frontal white matter apparent diffusion coefficients in patients with leukoaraiosis. Journal of Neurology, 255, 360-366.

What is Executive Function?

July 10th, 2009 5 comments

Executive function is a term that describes a wide range of cognitive behaviors and processes. It is broad enough of a term that some people simply describe it as, “what the frontal lobes do.” When asked what exactly the frontal lobes do do, some revert to the circular definition of “executive functions.” However, executive functions are distinct from – but related to – what the frontal lobes do. The frontal lobes are involved in motor functions (e.g., pre-motor and primary motor areas), eye movement (e.g., frontal eye fields), memory (e.g., acetylcholine-producing portions of the basal forebrain), and language (BA 44,45 or Broca’s area). In addition, some executive functions incorporate areas of the brain outside the frontal lobes – the parietal lobes or basal ganglia, for example. Like many cognitive domains, executive functions are part of a distributed network of brain structures and regions.

Most neuropsychologists however, would define (or at least accept the following definition of) executive function similar to this: Executive function is the ability to selectively attend to, work with, and plan for specific information. This means that executive function is deciding what information, cognitions, or stimuli are relevant, holding and working with that information, and then planning what to do with it. As such, executive function is largely the roles of planning and organization. It is also the ability to recognize and learn patterns (i.e., cognitive sets) but also have the cognitive flexibility to respond to set changes and make a shift in set. Executive function also involves being able to select the appropriate response or behavior while at the same time inhibiting inappropriate responses or behaviors.

Executive functions have been compared to the conductor of an orchestra who, in order to make sense of the disparate instruments, sounds, and parts, must coordinate the members and lead the efforts of all the components of the orchestra. Executive functions also have been compared to chief executive officers of companies. These comparisons demonstrate that executive functions are arguably the most complex and highest of all cognitive functions. However, just like most other cognitive functions, executive functions are comprised of relatively simple processes (e.g., attention and processing speed) – it is just the unique combination of these more basic processes that makes executive functions so powerful.

One potential problem with executive function as a cognitive domain is that it is large and loose. Many tests have been developed, or at least used, to assess executive function (e.g., Wisconsin Cart Sort Test, Stroop Color-Word Task, clock drawing, and so forth). Even though all such tests are used as measures of executive functioning, scores on them do not always correlate highly with each other. They do not always cluster together when subjected to principal components analysis or even structural equations modeling. This means that even though neuropsychologists have many purported tests of executive function, they all seem to measure different aspects of executive function. This might partially result from executive functioning tests being differentially affected by basic cognitive processes such as processing speed.

Even though, as previously mentioned, I do not believe executive functions and frontal lobe functions are synonymous terms, are we able to localize executive functions to the frontal lobes? Largely we can. The most evidence from neuroimaging studies and neurological injuries demonstrate that the prefrontal cortex – the area of the brain that is phylogenetically youngest and most advanced and as such, proportionately larger in humans than any animal – is necessary (but not necessarily sufficient) for executive functioning. When this area is disrupted in humans, they exhibit poor decision-making skills, including poor planning and poor maintenance or self-regulation of behavior. One area of the prefrontal cortex particularly involved in executive functions is the dorsolateral prefrontal cortex (area 46) – although both the orbitofrontal and anterior cingulate are involved in aspects of executive functions.

In 1986 Alexander, Delong, and Strick published their seminal work on five parallel and closed cortico-striato-thalamo-cortical loops. These frontal-subcortical circuits were hypothesized to be involved in a range of behaviors and cognitions based on the varying cortical connections of the loops. Previously, many researchers did not well-understand the role that the basal ganglia played in any sort of “higher” function; in fact, most viewed the basal ganglia as involved mainly in motor behaviors. Alexander, Delong, and Strick’s article set off a flurry of research into the functions of these frontal-subcortical circuits, which have been verified as existent in humans (Middleton & Strick, 2000). Over time different theories have modified these circuits, including that they are composed of direct, indirect, and hyperdirect pathways, which all function at different speeds or timings to allow the basal ganglia to regulate behavior. Mink (1996) proposed that actions (e.g., producing a specific word) are regulated by the direct and indirect pathways, which serve as large components of our ability to select and inhibit correct and incorrect responses, respectively. It is as if each individual fronto-cortical loop allows us to properly attend to the correct behavior or response and inhibit all other behaviors or responses, much like the DLPFC and orbitofrontal cortex and their associated loops are involved in the selection and inhibition of behavior, both major aspects of executive function.

Just as damage to the dorsolateral prefrontal cortex (DLPFC) produces deficits in executive function, damage to any part of the DLPFC loop also results in executive dysfunction. Benke, Delazer, Bartha, and Auer (2003) presented two clinical cases of patients with left caudate lesions (the lesions also affected part of the anterior limb of the internal capsule as well as portions of the putamen and pallidum; however, the infarcts affected the caudate the most). Among other deficits, both patients had executive function impairments, including problem-solving deficits, many perseverative errors, and set-shifting problems. Even though the patients had no direct DLPFC damage, they exhibited similar deficits to patients with DLPFC lesions. These executive deficits persisted over time.

As a cognitive domain, and even as broad as it might be, executive functioning has ecological validity. Price and colleagues (2008) found that executive dysfunction was related to greater difficulty performing IADLs. Specifically, patients with executive dysfunction had more difficulty performing IADLs than patients with memory deficits did. Thus, how quickly, flexibly, and accurately people can organize, solve, plan, or attend to specific neuropsychological tasks seems to correlate with their accomplishment of everyday tasks of life, such as finances, driving, and shopping.

References

Alexander, G. E., DeLong, M. R., & Strick, P. L. (1986). Parallel organization of functionally segregated circuits linking basal ganglia and cortex. Annual Review of Neuroscience, 9, 357-381.

Benke, T., Delazer, M., Bartha, L., Auer, A. (2003). Basal ganglia lesions and the theory of fronto-subcortical loops: Neuropsychological findings in two patients with left caudate lesions. Neurocase, 9, 70-85.

Middleton FA, & Strick PL. (2001). Basal ganglia output and cognition: evidence from anatomical, behavioral, and clinical studies. Brain Cogn., 42, 183-200.

Mink, J. W. (1996). The basal ganglia: Focused selection and inhibition of competing motor programs. Prog Neurobiol, 50, 381-425.

Price, C.C., Garvan, C., and Monk, T. (2008). Type and severity of cognitive impairment in older adults after non-cardiac surgery. Anesthesiology, 108, 8-17.

Optical Illusions That Make You Fatter and Your Wallet Lighter

October 30th, 2007 No comments

“Eat from small plates, drink from taller glasses.” Optical illusions lead us to eat and drink more, as illustrated by the examples in this article. There’s an old saying in cuisine…”the first bite is with the eye.”

Interesting article. I’m not sure if there is empirical data to support it but it does show that our perception of our food can affect how much we eat. Our actions are affected by so many different things, many of which we might not be aware.

read more | digg story

The Modal Model of Memory and the Serial Position Effect

July 8th, 2007 No comments

I’m continuing my recent trend of basic cognitive psychology posts. The following post is about the Modal Model of memory, which has been highly influential for a number of decades but it is slowly being modified over time. I won’t get into the more modern modifications of the modal model, rather, in my post I present the very traditional view of memory, even if it is somewhat controversial today. For example, a number of psychologists do not believe that short term memory really exists (working memory fills in the gap). In any case, my post serves as a brief introduction to a classic view of memory and of the primacy and recency effects.

The modal model of memory has three main components. They are: sensory register, short-term memory (STM), and long-term memory (LTM). This Atkinson and Shiffrin model of memory assumes that the processes of moving information from the sensory store to short-term and then long-term memory takes place in discrete stages. At any of these stages information can be lost through interference or decay. Another assumption of this model is that information processing has to start in the sensory register and be attended to, then move to STM, and then to LTM with rehearsal.

The serial position effect (split into the primacy and recency effects) is that the first few and last few items in a word list, for example, are the easiest to remember. A graph of this effect would be roughly parabolic (i.e., U-shaped). The primacy effect occurs because people have time to rehearse the first few items until the STM capacity is reached. The recency effect occurs because the last items are still in STM and have not decayed yet so they are easy to remember. The items in the middle of lists are easy to forget because STM capacity is too full for much rehearsal by then and as more items are presented, older items in STM are “pushed out.”

Serial Position EffectThere are ways to hinder the primacy or recency effects though. If items are presented rapidly then there is not time to rehearse the items and the primacy effect fades away. If there is a distracting task given at the end of the main task (similar to Peterson and Peterson’s 1959 study testing the decay rate of STM), then the recency effect disappears due to STM capacity being taken up by the distracters, which leads to decay of the information in STM. These findings indicate that the systems governing primacy and recency effects are separate. The findings also gave support to the modal model because researchers identified the primacy effect with the transfer of STM into LTM. The recency effect is just an example of information being in STM.

Word Superiority Effect and Parallel Processing

July 6th, 2007 2 comments

WordsOne experiment about cognitive brain functioning is the word superiority effect findings of Dr. Reicher in 1969. In this experiment either a word or a non-word (string of letters) is flashed on a screen. The subject is asked if the stimulus contained one of two letters, say a “C” or an “E”. When the stimulus did not resemble a word (e.g., XXCX) subjects were correct in identifying the target letter about 80% of the time. When the string of letters was similar to a word but not one (e.g., FELV) the subjects also correctly identified the target letter 80% of the time. However, the interesting finding was that when the stimulus was a word (e.g., TEND), subjects were correct in identification 90% of the time. So the word superiority effect is that subjects are most accurate in identifying a target letter when it is contained in a word as opposed to a string of letters.

This lends support to the theory that there are things that we can process in parallel and that that parallel processing (or parallel activation of word and letter) can be beneficial at times (such as helping subjects correctly identify individual letters more often when the letter is contained within a word rather than in a random string of letters). In other words, the whole word is recognized before all the letters individually are recognized. This then speeds up or aids processing because there are now a couple routes, per se, to that letter; there is the visual stimulus (seeing the letter) and the linguistic information (knowing that the letter is in the word) that both are activated and help people recognize letters better.

Image by uncommonmuse.