Are effect sizes and confidence levels problems for or solutions to the null hypothesis test?

Some have proposed that the null hypothesis significance test, as usually conducted using the t test of the difference between means, is an impediment to progress in psychology. To improve its prospects, using Neyman-Pearson confidence intervals and Cohen's standardized effect sizes, d, is recommended. The purpose of these approaches is to enable us to understand what can appropriately be said about the distances between the means and their reliability. Others have written extensively that these recommended strategies are highly interrelated and use identical information. This essay was written to remind us that the t test, based on the sample--not the true--standard deviation, does not apply solely to distance between means. The t test pertains to a much more ambiguous specification: the difference between samples, including sampling variations of the standard deviation.

Treatment of the cutaneous pain of acute herpes zoster with 9% lidocaine (base) in petrolatum/paraffin ointment.

BACKGROUND: Treatment of the pain of acute herpes zoster by local anesthetic injections has drawbacks. Topical percutaneous local anesthesia (TPLA) may offer another strategy of providing regional analgesia in affected patients. OBJECTIVE: We evaluate the analgesic efficacy and safety of 9% (wt/vol) lidocaine (base) in petrolatum/paraffin ointment in patients with acute herpes zoster. METHODS: Ointment was applied to the affected skin of 22 patients. Pain, tenderness, sensitivity to pinprick and cold, and blood lidocaine concentration were measured repeatedly during a 20-hour interval and intermittently thereafter. RESULTS: Mean pain, tenderness, and cutaneous sensation scores were reduced at measurements taken from 4 to 20 hours after ointment application (p < 0.05), but not every patient obtained relief. No patient had local skin irritation or systemic toxic effects related to the local anesthetic. CONCLUSIONS: TPLA is a promising therapy for control of cutaneous pain of acute herpes zoster. Controlled studies should be performed to prove efficacy, determine optimal TPLA formulation, and define dosage limits.

Postnatal protein deprivation in rhesus monkeys.

We studied the growth of 97 rhesus monkeys (M. mulatta) that from birth until 120 days of age were fed a diet containing 13.4%, 6.7%, 3.35% protein or a commercial simulated human-milk formula (SMA) affording 9% protein. After 120 days, all animals were fed SMA. Females fed the diet lowest in protein, but not the other diets, were moderately affected, gaining less weight than their well-fed counterparts did, but they recovered their deficit during the repletion period, so that at 240 days no group differences remained. Males fed the two lowest-protein diets were severely and profoundly affected, in keeping with the depth of their protein deficiency. Moreover, unlike the females, they recovered none of their weight deficit during the repletion period. Food intake during deprivation was lower in the animals eating the low-protein diets. During SMA repletion, intakes followed no simple rule, but they did not converge. Except for the males fed the 3.35% protein diet, relative food intake (ml formula/kg body weight) did not differ substantially between diets or sexes at any time during the deprivation and diminished as the animals got older. Those males' relative intakes did not decline. During repletion, the SMA females ate the most in proportion to weight and the 13.4% group the least. The lowest-protein males, in contrast, ate the most in proportion to their weight during both deprivation and repletion. Males fed the lowest-protein diet gained little for what they ate; those fed the highest-protein diet gained much. Females were more efficient than males were when eating the low-protein diet and less efficient when eating the high-protein diet. When fed SMA during repletion, males' food efficiencies (grams gained/liter of diet) were nearly equal; females still differed: the SMA group was the least efficient and the low-protein group most efficient. Its counterpart, protein efficiency, was greatest (during deprivation) for females eating the low-protein diet and least (among females) for those eating the high-protein diet. Males were least protein efficient if eating the low-protein diet. Evidently, a 4 month bout of protein deprivation had prolonged effects on the amount of food the animals needed to produce a given gain in weight.

Intraday feeding patterns in infant rhesus monkeys and the effects of missing a meal.

During the first month of life, infant rhesus monkeys (N = 96) were fed 7 times a day for the first 21 days and were fed 6 times a day thereafter. Fifty monkeys were fed SMA, a formula designed for human infants (9% protein, 43% carbohydrate, and 48% fat); 46 were fed one of three laboratory-confected diets varying in the amount of protein and carbohydrates provided. Although the diets had differential effects on weight gain, overall daily consumption was practically unaffected. A characteristic pattern was evident soon after birth: The first meal of the day (8:00 AM) was one of the largest, and the following two meals were usually the smallest; the next peak spanned the 4:00 PM and 8:00 PM feedings. The meal following a missed meal was slightly larger than normal. To some extent, the large meal at 8:00 AM can be attributed to the 10-hr absence of food during the night, but the peak at 4:00 PM and 8:00 PM cannot be so accounted for. The adult bicuspid pattern of eating hence appears to develop very early in life.

Probability-reward preferences of rhesus monkeys.

The relative preferences of rhesus monkeys for reward probability versus amount were investigated with procedures which contrasted general experience with specific instructions, and evaluated the relationship between probability-amount combinations and preference strength. Four stimulus objects, each signifying a different combination of reward frequency and amount (100% with one unit; 50% with two units; 33% with three units; or 25% with four units), were presented in pairs, one pair per daily session, with trial schedules providing the same amount of reward within each set of 12 trials. In Phase A, 4 monkeys (Group 1) were tested on the six choice-pairs with no preliminary training. In Phase B, Group 1 was joined by an additional 4 monkeys (Group 2), and each of the tasks was preceded by a demonstration of the relevant stimuli, one at a time, together with their associated probabilities and amounts. Group 1 animals developed preferences during Phase A for the more frequently rewarded objects, which persisted into Phase B, whereas Group 2 animals showed no preferences. This result indicates that preliminary instructions concerning the reward combinations associated with stimulus objects can prevent the development of a preference for greater probability over greater amount of reward, but cannot extinguish it once it has been formed or reestablished within the context of a particular task.

Scotopic thresholds of rhesus monkeys at different grating densities under normal and low-protein diets.

Scotopic thresholds of rhesus monkeys were determined over a range of luminance levels, from 1138 to 17.8 x 10(-9) cd/cm2, and grating densities, from .132 to 1.58 lines per cm. The effects of a low-protein diet on these thresholds were also investigated, and standard stimulus objects within the discrimination learning situation were employed. Obtained thresholds decreased systematically from about 350 to 20 x 10(-9) cd/cm2, and there were no significant differences due to dietary protein level.

Loci of discriminanda and discrimination proficiency.

Three groups of nine juvenile rhesus monkeys (Macaca mulatta) whose mothers were maintained on diets affording 3.35, 6.70, and 13.4 percent of calories as protein during gestation were tested in a series of discrimination-learning tasks in which the critical variable was the loci of the relevant cues. The stimuli used were two 75-mm x 75-mm plastic plaques. The loci of the discriminanda (colored dots on the plaque) varied according to a 3 x 3 grid. No significant differences were found among the diet groups. A gradient of discriminability appeared that accorded with the location of the discriminanda: The stimulus region closest to the monkey made for the least and the farthest region the most difficulty of discrimination, also discriminanda on the right part of the stimulus were easier to discriminate than those on the left. Animals rewarded for displacing the plaque containing a colored dot learned the discriminations faster than those reinforced for displacing the blank plaque. The differences in salience of different parts of a stimulus plaque, though systematic, are modest, not dramatic.

Vesicoureteral reflux in the primate: III. Effect of urinary tract infection on maturation of the ureterovesical junction.

Recurrent bladder infections in the monkey appeared to cause a delay in maturation of the ureterovesical junction. This was not significant since it did not ultimately prevent maturation. Bacteriuria was prolonged when reflux was present.

Protein deprivation in primates. X. Test performance of juveniles born of deprived mothers.

Forty juvenile rhesus monkeys whose mothers were fed semisynthetic diets containing 13.4, 6.7, or 3.35% protein during pregnancy were tested as juveniles on four tests of perceptual-discrimination learning. All juveniles were well fed since birth. No significant differences in proficiency due to diet could be detected. It is concluded that thanks to the mother's proficient adaptation to the diet the fetus is well protected, and not deprived.

Protein deprivation in primates. XI. Determinants of weight change during and after pregnancy.

Individual differences in weight gain of rhesus monkeys during pregnancy far exceed the variation in infant weight. Weight gain thus reflects maternal adjustment to her physiological state in relation to her environment rather than growth of the infant. Diets high in protein lead to great weight gain. Skeletal length is positively correlated and conception weight negatively correlated with weight gain. Season of pregnancy, length of gestation and sex of the infant are unrelated to weight gain. None of the monkeys exhibited edema during gestation. Postpartum changes in body weight were small during the first 6 weeks.

Protein deprivation in primates: XII. food preferences of juvenile rhesus monkeys.

21 male and 18 female juvenile rhesus monkeys whose mothers' prenatal diets varied according to protein content (either 1-, 2-, or 4-gm protein/kg body weight/day) were given a food-preference task. 20 monkeys were between 3 and 4 yr. old; 19 were between 1 1/2 and 2 1/2 yr. old. For each trial small pieces of the same type food were placed in 8 equally spaced wells about the circumference of a circular tray held outside each monkey's cage. Responsiveness to the 21 natural foods (assorted cheeses, meats, vegetables, nuts, cereals, fruits, and candies) offered 3 times on 3 different occasions was measured by recording number of pieces taken, number eaten, and elapsed time for taking all 8 pieces. All 3 measures were highly correlated. All diet groups exhibited the same general order of preference, males taking and eating more food pieces faster than females. Order of preference was significantly different for old versus young juveniles, with older juveniles' preferences more similar to those of adult female monkeys given the same task. The most notable differences in eating behavior were that younger juveniles played with their food and were less finicky about what they ate. Results suggest that food preferences of juvenile monkeys are not altered as a consequence of their mothers' prenatal protein deprivation.

Vesicoureteral reflux in the primate: II Maturation of the ureterovesical junction.

Infant rhesus monkeys frequently show vesicoureteral reflux unlike adults. We studied monkeys born of mothers who had been on normal or protein-deficient diets. Protein deficiency may influence the time that reflux continues, but this was not statistically proved. Reflux disappears over time. The time was influenced by when the pregnancy was initiated and the maturity of the infant. Maturation as measured by skeletal maturity scores correlated well with the time of reflux, but growth as measured by birth weight did not. Vesicoureteral reflux in the monkey is frequent, but maturation of the ureterovesical junction occurs by age 3 years and reflux disappears.

Protein deprivation in primates. VIII. Early behavior of progeny.

Infant rhesus monkeys whose mothers were fed diets providing 1, 2, or 4 g of protein/kg body weight/day during pregnancy were tested on 13 behavioral tests during their first 60 days of life. (The diets had been shown to affect significantly the weights and blood protein concentrations of both pregnant and nonpregnant females.) The tests samples a wide range of behaviors, including visual and auditory perception, feeding reflexes, and locomotion. Despite the maternal dietary differences, the performances of infants on all tests were comparable in all groups. These data contrast with the finding of persistent deficit in rats born of protein-deprived mothers.