Education and demonstration – Psychology
Reexamination Certificate
2001-05-22
2004-03-30
Harrison, Jessica (Department: 3714)
Education and demonstration
Psychology
C434S238000, C434S323000, C434S350000, C434S428000
Reexamination Certificate
active
06712615
ABSTRACT:
TECHNICAL FIELD
The present invention relates generally to a system and method for internet-based cognitive performance testing.
BACKGROUND OF THE INVENTION
Systems and methods for computer based testing (CBT) are known to the art. For example U.S. Pat. No. 5,827,070 to Kershaw et al. discloses CBT means for administration of standardized test, e.g. SATs, LSAT, GMATs, etc. The system and method of Kershaw et al. does not depend on the speed and accuracy of the individual examinee's keystroke responses to the test stimuli. Lewis et al. (U.S. Pat. No. 5,059,127) disclose a computerized mastery testing system providing for the computerized implementation of sequential testing. This disclosure also does not relate to the speed and accuracy of the individual examinee keystroke responses to the test stimuli. Swanson et al. (U.S. Pat. No. 5,657,256) disclose a method and apparatus for administration of computerized adaptive tests. Swanson is similarly unconcerned about the examinee response time.
When people measure their response time, recall or other cognitive skills using computer-based test systems, they typically press number or letter keys (or keys representing other symbols like circles and squares) in response to visual or auditory or other sensory signals presented to them. The average time they take to press the correct keys is their response time.
This type of measurement is subject to a number of errors that make response time results relatively imprecise. The effects of recent foods and beverages, medicines, amount of sleep and other factors that affect alertness or drowsiness all influence response speed and accuracy, so that measurements on any single day may not represent actual average performance level.
A key source of measurement error is change in motivation to respond quickly. One day a person may try quite hard to reduce their response speed. The next day they may relax and perform more slowly simply because they care less about their “score” that day. Typically the error rate increases (incorrect responses are made more frequently) when people try harder to react quickly. Investigators commonly measure error rate to determine the “response speed/accuracy tradeoff” for each person or group of people.
While the response speed/accuracy tradeoff is usually discussed in connection with relatively simple responses, a similar tradeoff can occur during memory measurements when speed is only a secondary consideration. Response speed is intrinsically linked with recall accuracy because transient memory traces fade if the response (e.g. typing a list of words) is not completed rapidly.
Response speed may also vary from second to second and minute to minute as a result of boredom with the test, short-term fatigue from repeated motion, eye strain from staring at the computer screen, and stimulus patterns that confuse the user and cause response errors. Different types of transitions such as shifts between responses involving one hand and the other, or one finger and the corresponding finger on the other hand, can also affect response speed and accuracy for individual responses.
All of these factors together make precise performance measurement all but impossible. Even under controlled laboratory conditions, the correlation between test scores on one occasion and scores by the same individuals at a later time, averages only 0.63 (Salthouse & Babcock, 1991; Lowe & Rabbitt, 1998; Versavel et al., 1997; Wetherell, 1996). In other words, performance results can vary by plus or minus 20% from one day or week to the next.
The correlation between test results on separate days, called “test-retest reliability,” is perhaps the most widely used indicator of measurement reliability. The average value of 0.63 has not changed appreciably during the last two decades, indicating that attempts to improve measurement reliability have generally met with little success.
Perhaps the best way to describe the need for measurement precision, and the need for this invention, is to discuss the circumstances of an individual who participated in a recent study to determine whether blueberries can reduce multiple sclerosis symptoms (Pappas et al., 2001).
SF is one of hundreds of thousands of people in the U.S. who have chronic, neurodegenerative diseases for which there is no cure. He cannot drive and cannot find work because his coordination and memory are affected. He must sell the home he, his wife and children live in because they need the money for his medical and dental expenses. His relatively expensive medicines give no apparent benefit. The medicines do however dry his mouth and cause his teeth to crack, causing him to lose three teeth during the last several months. Concerned about his dental bills, SF asked his dentist to remove his remaining teeth so he would not have to pay to have them repaired when he would lose them anyway. (His dentist refused.) His physician advised him to take a recommended performance test battery just once a year because he cannot afford the cost of more frequent evaluations. He must therefore wait for very long periods of time before obtaining objective evidence that his medications are or are not helping him—time he can ill afford since his disease is growing steadily worse. And of course after such long waiting periods, any performance benefits provided by his medications may be cancelled by the steady decline from his chronic illness.
SF can expect to decline at a rate which reduces his performance scores by roughly 4% to 10% each year. If his medicines are effective, his annual decline may be decreased by half a percentage point or perhaps several percentage points—however he most probably cannot measure this benefit because once-a-year testing is not accurate enough to measure changes smaller than 5%. Once-a-year testing will always be incapable of measuring changes of 5% or less simply because he may perform 5% or 10% better or worse than his average on the day when measurement is performed.
So the test results for which SF must wait so long, and pay so much for, are largely worthless to him and his physician since they will not be precise enough to indicate whether his medicines helped him.
SF clearly needs, and many thousands of other people in similar circumstances need, a test system that is accurate to within 1% or 2% so that effective treatments can be identified. He also needs a measurement system that is far less expensive than that recommended by his physician, so that he can obtain results many times each year. And he needs a test that can be taken at home, so that he is spared the effort and/or the cost of transportation to a test center.
For these and many other reasons, there is a clear need for increased measurement precision
One strategy used by scientists seeking greater precision is to reduce response time variability by discarding high and low responses within each test or test series. For example, the slowest half and the fastest quarter of response times may be discarded from each 30 seconds of testing, and the average of the remaining data obtained.
This type of data trimming certainly reduces variability—but it also reduces the amount of useable data and therefore reduces measurement precision, which is related to the amount of data. (As a general rule, precision is directly proportional to the square root of the number of data points, if approximately random variation is the cause of imprecision.)
Discarding high response times also prevents or sharply reduces the accuracy with which benefits or harm from different health strategies can be measured if performance changes occur primarily within the response times that are discarded. This occurred recently during an Danbury MS Blueberry Study (Pappas et al, 2001). Very slow response times were markedly reduced after blueberry consumption for many study participants, however this was not evident from trimmed data sets, from which all slow responses had been removed. Only when raw data was examined did the principal investigator see this benefit.
Scientists have also attempted to red
Connolly Bove & Lodge & Hutz LLP
Evans John A.
Harrison Jessica
Sotomayer John
LandOfFree
High-precision cognitive performance test battery suitable... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with High-precision cognitive performance test battery suitable..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and High-precision cognitive performance test battery suitable... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3224954