Notes

1. Giving examples of categories: Eight common categories like CLOTHING with a large number of concrete instances were used. Normative data were available on

Constrained

90

80

9T

70

o a>

60

o

O

50

40

X Iii

T"

Breath Holding Duration

Fig. 1. Acute effects of marijuana on paired-associate learning. Significant impairments produced by marijuana containing 19 mg of THC (relative to placebo) and long (relative to short) breath-holding duration, and drug x associative type and drug x breath-holding duration x associative type interactions, for percentage correct. The research assistant guided the subjects in paced smoking procedures, which involved long or short breath-holding durations for different subjects. The drug x breath-holding duration interaction is shown separately for each of five associative types of constrained associations (left) and three associative types of free associations (right) (see Note 9). The drug x associative type and drug x breath-holding duration x associative type interactions were attributable solely to constrained associations. The "example" associative type was primarily responsible for the drug x associative type interaction for constrained associations. The overall impairment in learning produced by marijuana was shown by all the other associative types, but was reversed for the "example" associative type. The drug x breath-holding duration x associative type interaction for constrained associations partly reflected greater marijuana effects when the breath-holding duration was long rather than short for three associative types ("category," "property," and "another"), but not the others. The error bars indicate standard errors. Abbreviations: ANOTH. = another; CAT. = category; EXAMP. = example; L = long breath-holding duration; MOD. = moderately strong free associates; OPP. = opposite; PROP. = property; S = short breath-holding duration; UNASS. = unassociated words; WEAK = weak free associates. (From ref. 5.)

the frequency with which various instances are given as examples of the categories (10). Each subject was tested individually. The subject was instructed to give as many instances of each category as he or she could, in whatever order they came to mind. The experimenter spoke a category name and the subject gave instances

CLTR CLTR LTR LTR STR STR High Low High Low High Low

Score and Imagery

Fig. 2. Effects of chronic marijuana use on retrieval in Buschke's Test. Significant user group x imagery interactions for consistent long-term retrieval, long-term retrieval, and short-term retrieval scores. The chronic marijuana users were classified as heavy, intermediate, or light users. These groups were defined by usage seven or more times, five to six times, or one to four times weekly on average for the previous 2 yr or more, respectively; and are represented by black bars, white bars, and bars with diagonal lines, respectively. Nonusers are represented by bars with horizontal lines. The scores, which are averaged over trials, are based on eight high-imagery words and eight low-imagery words (see Note 10). The error bars indicate standard errors. Heavy users showed impairment relative to nonusers in long-term retrieval for high-imagery but not low-imagery words (center). Consistent long-term retrieval, which measured the numbers of words reliably recalled from trial to trial without omission, showed a similar pattern, i.e., impairment in heavy users relative to nonusers for high-imagery but not low-imagery words (left). The data for short-term retrieval suggested a slight, opposite, compensatory tendency, but were equivocal (right); although the user group x imagery interaction was significant, nonusers did not differ from heavy users (or from intermediate or light users). Significance is indicated as follows: * heavy users differed from nonusers by Dunnett's test, p < 0.05. Abbreviations are as follows: CLTR = consistent long-term retrieval; High = high-imagery words; Low = low-imagery words; LTR = long-term retrieval; STR = short-term retrieval. (From ref. 6.)

1 2 Session

Fig. 3. Effects of chronic marijuana use on presentations to criterion in Buschke's Test. Learning a list of words to a criterion of two consecutive perfect recalls during an initial session, followed by relearning of the same list, again to a criterion of two consecutive perfect recalls, during a subsequent session (see Note 11). Numbers of presentations of the list required to reach the criterion are shown. Abbreviations: 1 = initial session; 2 = subsequent session. Chronic marijuana users performed more poorly than control subjects (nonusers), i.e., they required more presentations in learning and relearning the list to criterion during these two sessions. Not surprisingly, far fewer presentations were necessary to relearn the list during the second session than to learn it initially during the first session. Despite this, the impairment in the chronic marijuana users did not vary between sessions, i.e., the number of extra presentations required by chronic marijuana users, relative to control subjects, was similar for initial learning and relearning, F < 1 for the interaction of group and session. (From ref. 7.)

for 2 min. The responses were tape-recorded. Subjects rested for 1 min between categories. The order in which the categories were presented was counterbalanced over subjects. In scoring, exact repetitions of instances given previously were deleted. Subsequently, ratings of typicality were obtained for all instances given by subjects from a separate group of college students.

2. Giving examples of categories beginning with specified letters: On each of a series of trials, subjects were presented with a category name followed by a letter (e.g., WEAPON — G), and had to name an instance of that category beginning with that letter (e.g., GUN). The "target" for each category-letter combination was defined as the most frequently given instance of the category beginning with the specified letter according to frequency norms (10-11). Some trials used category-letter combinations with "common targets," while others had "uncommon targets." For

1 2 Session

Fig. 3. Effects of chronic marijuana use on presentations to criterion in Buschke's Test. Learning a list of words to a criterion of two consecutive perfect recalls during an initial session, followed by relearning of the same list, again to a criterion of two consecutive perfect recalls, during a subsequent session (see Note 11). Numbers of presentations of the list required to reach the criterion are shown. Abbreviations: 1 = initial session; 2 = subsequent session. Chronic marijuana users performed more poorly than control subjects (nonusers), i.e., they required more presentations in learning and relearning the list to criterion during these two sessions. Not surprisingly, far fewer presentations were necessary to relearn the list during the second session than to learn it initially during the first session. Despite this, the impairment in the chronic marijuana users did not vary between sessions, i.e., the number of extra presentations required by chronic marijuana users, relative to control subjects, was similar for initial learning and relearning, F < 1 for the interaction of group and session. (From ref. 7.)

instance, WEAPON — G had a common target (GUN), whereas WEAPON — B had an uncommon target (BOMB), because when people give examples of WEAPONS, GUN is a more common example than BOMB. The experiment provided three measures for analysis: "response rate," the percentage of trials in which any response was given; the RT for those responses; and "percent of targets," the percentage of the subject's responses that were the expected targets. Each category was used on two trials with different letters (e.g., WEAPON — G, then WEAPON — B). One third of the categories were tested at "long lags" and the remainder at "short lags," the lag being defined as the number of trials on other, unrelated categories that intervened between the two trials on a given category. A series of 126 trials was formed using 63 categories and two letters for each category. The experimenter spoke a category name, following which a letter was presented by a tachistoscope. The subject responded orally. RT was measured by a voice-operated relay.

3. Free and constrained associations. For each of 100 words presented at a 10-s rate on a computer monitor, the subject gave a single word as an association. Response times were measured by a voice-activated relay. For 50 words, the subject was cued to provide a "free association," i.e., any kind of association that came to mind; for 10 words each, the subject was to provide one of five types of "constrained associations," i.e., a response that related to the stimulus in a specified way, cued by the words "another," "category," "example," "opposite," or "property." Examples of these types are "book-magazine," "aluminum-metal," "fruit-apple," "night-day," and "banana-yellow," respectively. The words were drawn from those used in a pilot study (12) involving other subjects, so that for each response in the present study the number of pilot subjects giving that response ("dominance") could be scored, e.g., eight pilot subjects gave "sandal" in response to "shoe."

4. Deciding if item belongs to specified category: Subjects saw a category name followed by an instance, and decided whether the instance was an example of the category as fast as possible. The correct response was "yes" for half the trials (e.g., FRUIT — APPLE) and "no" for the others (e.g., FRUIT — CROWN). The instance was a common example of the category on half of the "yes" trials, an uncommon example on the other half. Categories and instances were selected using normative data (10). Each of 32 categories appeared on two "yes" trials (one with a common instance and one with an uncommon instance) and on two "no" trials (with two different noninstances, i.e., items belonging to other categories). For each trial the category and, 2 s later, the instance were presented by a tachistoscope. The subject responded by pressing one of two switches labeled YES and NO to indicate whether or not the instance belong to the specified category. RT was measured.

5. Deciding if two items belong to same category: Subjects viewed two instances and decided whether they belonged to the same category (e.g., APPLE PEACH) or to different categories (e.g., APPLE BLUEBIRD). Half the trials were primed, half unprimed. On primed trials only, before seeing the two instances, subjects heard a prime, i.e., a category name like FRUIT, which informed them that one or both of the forthcoming instances would come from that category. Half of the primed and unprimed trials involved common examples of categories; the other trials involved uncommon examples. For each combination of priming condition and type of instance (common vs uncommon), the correct response was "same" for half the trials and "different" for the rest, and the "same" trials were evenly divided between "same category" trials like APPLE PEACH, involving two distinct instances from the same category; and "same instance" trials like APPLE APPLE, on which a single instance appeared twice. To form 80 primed trials, 10 categories were chosen. Each category was used as the prime on eight trials, which were allocated in the required proportions to "same category," "same instance," and "different" trials with common vs uncommon instances. Each specific instance was used in only one primed trial. For each primed trial formed in this way, there was a corresponding unprimed trial using the identical pair of instances. On each trial, the experimenter spoke the category name on primed trials or the word "blank" (as a warning signal) on unprimed trials, following which the pair of instances was presented by a tachistoscope. The subject responded by pressing one of two switches labeled SAME and DIFFERENT to indicate whether or not the two instances belonged to the same category. RT was measured.

6. Letter-matching: On each trial, subjects saw two letters and decided whether or not they had the same name. Three kinds of trials were mixed together: "same-case" trials like "AA," on which the letters were physically identical; "same-name" trials like "Aa," on which the letters had the same name but appeared in opposite cases; and "different" trials like "Ab," on which the letters had different names. There were 80 "primed" trials and 80 "unprimed" trials. On primed trials only, before seeing the pair of letters, subjects heard a prime (i.e., a letter name like "A"), which informed them that one or both of the forthcoming letters would have that name (e.g., would be "A" or "a"). For each priming condition, the correct response was "same" for half the trials and "different" for the remainder, and the "same" trials were evenly divided between "same case" and "same name" trials. Ten letter names were used with equal frequencies as the prime on primed trials. The experimenter spoke a letter name on primed trials or the word "blank" (as a warning signal) on unprimed trials, following which a pair of letters was presented by a tachistoscope. The subject responded by pressing one of two switches labeled SAME and DIFFERENT to indicate whether the two letters had the same name. RT was measured.

7. Paired-associate learning: Three lists of high-imagery nouns, each consisting of 20 pairs of words, were tape-recorded. For each list, the tape contained two study trials (the 20 word pairs read sequentially) and two test trials (the 20 initial words read sequentially, with subjects instructed to orally supply the second word in each pair). The interval between successive pairs was 8 s. All subjects were given standard paired-associate learning instructions. Half the subjects were instructed to use visual imagery during learning—to try to learn by forming a vivid, detailed visual image of the two items in each pair in some interactive relationship. The other subjects were not instructed in any specific learning technique. Each subject received the two study and test trials on one list, then two on a second list, then two on a third list, with a rest between lists. After the testing, subjects who were given imagery instructions were asked to describe their images for each word pair. Subsequently, a separate group of college students rated the vividness of the described images.

8. Text learning: The subject read a paragraph from an article in Reader's Digest on a computer monitor at his or her own pace, pressed a button when finished, and then recalled as much as possible in 3 min. The subject then reread the paragraph and recalled it again. Following this, the entire procedure was repeated with a different paragraph. In addition, delayed recall was tested 15 min before the end of the session. The subject's recall was tape-recorded for later scoring of the propositions recalled (13). Reading times were determined based on the subject's button presses.

9. Paired-associate learning of associated and unassociated words: A list of 30 pairs of words was presented at a 3-s rate on a computer monitor. Following this study trial, the initial word from each pair was presented in a test trial at a 5-s rate (with subjects instructed to orally supply the second word in each pair). Response times were measured by a voice-activated relay. These study and test trials were then repeated. Following this, two study and test trials were administered on a second list, consisting of different pairs of words. Each list consisted of three pairs representing each of the five types of constrained associations used in the free and constrained associations test (see Note 3), mixed with 15 other pairs, of which equal numbers involved moderately strong free associates, weak free associates, and unassociated words, e.g., "plumber-pipe," "tell-secret," and "carpet-laughter," respectively (12).

10. Buschke's Test. A list of 16 nouns, half "high-imagery" words that were easy to visualize (e.g., "bouquet") and the remainder "low-imagery" words that were difficult to visualize (e.g., "replacement") (14), was presented on a computer monitor at a rate of 3 s per word. The subject tried to recall as many words as possible. Seven learning and test trials were given consecutively. The subject tried to recall the whole list on each test trial, but, on learning trials after the first, was reminded only of the words missed on the immediately preceding test trial. This "selective reminding" technique allowed scoring several aspects of memory in addition to total recall (9). In essence, recall of a word without an immediately preceding reminder (i.e., recall on two successive trials) indicated that the word had entered (and presumably remained in) long-term storage. Recall of the word before this occurrence was attributed to short-term retrieval; after it, to long-term retrieval. Long-term retrieval was designated "consistent" when the word was never subsequently omitted.

11. Buschke's Test and subsequent free recall. The subject learned a list of 15 common words, such as "drum," "curtain," etc., to a criterion of two consecutive perfect recalls, using Buschke's "selective reminding" technique (9) (see Note 10). The subject's recall was oral. In a subsequent session, the subject relearned the same list, again to a criterion of two consecutive perfect recalls. Next, the subject was played a recording of the list and recalled it according to slightly different proce dures to be used in a functional neuroimaging session on the following day (e.g., computerized presentation of a digitized recording of the list), and again relearned it. During the neuroimaging session, the subject tried to orally recall the list relearned on the previous day, without any prompting. The subject then heard the computerized presentation of this list again and immediately tried to recall it. Subsequently, the subject heard a list of 15 different common words and immediately tried to recall them. These three tests differed in their relative demands on episodic memory encoding and retrieval. During each test, the subject recalled the words for 40 s and was asked to repeat them if he or she could not think of any more. The two lists were from the Rey Auditory Verbal Learning Test (15).

Was this article helpful?

0 0

Post a comment