Conventional Levels Of Statistical Significance Research Articles

The data monitoring experience in the Candesartan in Heart Failure Assessment of Reduction in Mortality and morbidity (CHARM) program

The CHARM program was designed as 3 separate randomized trials comparing candesartan with placebo in patients with chronic heart failure (CHF) who (1) were intolerant to angiotensin-converting enzyme inhibitor and had left ventricular ejection fraction (LVEF) < or =0.40; (2) were on angiotensin-converting enzyme inhibitor and had LVEF < or =0.40; or (3) had LVEF >0.40. CHARM provides an interesting example of the challenges faced by a Data and Safety Monitoring Committee (DSMC). Although the primary efficacy end point for each component trial was cardiovascular (CV) death or hospitalization for CHF, the primary outcome for the overall program was all-cause mortality. The DSMC received monthly safety reports and also met every 6 months (7 times in all) to review interim reports. Statistical stopping guidelines were predefined for all-cause mortality in the overall program. The overarching principle of the DSMC was proof beyond a reasonable doubt that would be likely to influence clinical practice. There were significant treatment differences in all-cause mortality for the overall program at several interim analyses, and the statistical stopping guideline was reached on one occasion. However, even a conventional level of statistical significance (P < .050) was achieved in only 1 of the 3 component trials. The DSMC consistently recommended that the program continue as planned. The final published result for all-cause death over a median of 3.1 years was a 9% reduction in hazard (95% CI 0%-17%, P = .055), whereas for CV death or hospitalization for CHF, there was a 16% reduction in hazard (95% CI 9%-23%, P < .0001). Subsequent exploratory analyses suggest that the hazard reduction in CV death was more marked in the first year after randomization and that, if real, this apparent treatment-time interaction offers a plausible explanation for why the interim mortality data showed statistically more extreme findings than the overall final results. The DSMC experience in the CHARM program illustrates the importance of continuing a trial to its scheduled completion unless there is proof beyond a reasonable doubt that would influence clinical practice rather than strict reliance on a statistical stopping guideline.

Olanzapine for schizophrenia

Olanzapine is an atypical antipsychotic reported to be effective without producing disabling extrapyramidal adverse effects associated with older, typical antipsychotic drugs. To determine the clinical effects and safety of olanzapine compared with placebo, typical and other atypical antipsychotic drugs for schizophrenia and schizophreniform psychoses. We updated the first search [Biological Abstracts (1980-1999), The Cochrane Library (Issue 2, 1999), EMBASE (1980-1999), MEDLINE (1966-1999), PsycLIT (1974-1999) and The Cochrane Schizophrenia Group's Register (October 2000)] in October 2004 using the Cochrane Schizophrenia's Group's register of trials. We also searched references of all included studies for further trials, and contacted relevant pharmaceutical companies and authors. We included all randomised clinical trials comparing olanzapine with placebo or any antipsychotic treatment for people with schizophrenia or schizophreniform psychoses. We independently extracted data and, for homogeneous dichotomous data, calculated the random effects relative risk (RR), the 95% confidence intervals (CI) and the number needed to treat (NNT) on an intention-to-treat basis. For continuous data we calculated weighted mean differences. Fifty five trials are included (total n>10000 people with schizophrenia). Attrition from olanzapine versus placebo studies was >50% by six weeks, leaving interpretation of results problematic. Olanzapine appeared superior to placebo at six weeks for the outcome of 'no important clinical response' (any dose, 2 RCTs n=418, RR 0.88 CI 0.8 to 0.1, NNT 8 CI 5 to 27). Although dizziness and dry mouth were reported more frequently in the olanzapine-treated group, this did not reach statistical significance. The olanzapine group gained more weight. When compared with typical antipsychotic drugs, data from several small trials are incomplete. With high attrition in both groups (14 RCTs, n=3344, 38% attrition by six weeks, RR 0.81 CI 0.65 to 1.02) the assumptions included in all data are considerable. For the short term outcome of 'no important clinical response', olanzapine seems as effective as typical antipsychotics (4 RCTs, n=2778, RR 0.90 CI 0.76 to 1.06). People allocated olanzapine experienced fewer extrapyramidal adverse effects than those given typical antipsychotics. Weight change data for the short term are not statistically significant but results between three to 12 months suggest a clinically important average gain of four kilograms for people given olanzapine (4 RCTs, n=186, WMD 4.62, CI 0.6 to 8.64). Twenty three percent of people in trials of olanzapine and other atypical drugs left by eight weeks; 48% by three to12 months (11 RCTs, n=1847, RR 0.91 CI 0.82 to 1.00). There is little to choose between the atypicals, although olanzapine may cause fewer extrapyramidal adverse effects than other drugs in this category. Olanzapine produces more weight gain than other atypicals with some differences reaching conventional levels of statistical significance (1 RCT, n=980, RR gain at 2 years 1.73 CI 1.49 to 2.00, NNH 5 CI 4 to 7). There are very few data for people with first episode illness (1 RCT, duration 6 weeks, n=42). For people with treatment-resistant illness there were no clear differences between olanzapine and clozapine (4 RCTs, n=457). The large proportion of participants leaving studies early in these trials makes it difficult to draw firm conclusions on olanzapine's clinical effects. For people with schizophrenia it may offer antipsychotic efficacy with fewer extrapyramidal adverse effects than typical drugs, but more weight gain. There is a need for further large, long-term randomised trials with more comprehensive data.

Phase III pilot study of dose escalation using conformal radiotherapy in prostate cancer: PSA control and side effects

Radical radiotherapy is a standard form of management of localised prostate cancer. Conformal treatment planning spares adjacent normal tissues reducing treatment-related side effects and may permit safe dose escalation. We have tested the effects on tumour control and side effects of escalating radiotherapy dose and investigated the appropriate target volume margin. After an initial 3–6 month period of androgen suppression, 126 men were randomised and treated with radiotherapy using a 2 by 2 factorial trial design. The initial radiotherapy tumour target volume included the prostate and base of seminal vesicles (SV) or complete SV depending on SV involvement risk. Treatments were randomised to deliver a dose of 64 Gy with either a 1.0 or 1.5 cm margin around the tumour volume (1.0 and 1.5 cm margin groups) and also to treat either with or without a 10 Gy boost to the prostate alone with no additional margin (64 and 74 Gy groups). Tumour control was monitored by prostate-specific antigen (PSA) testing and clinical examination with additional tests as appropriate. Acute and late side effects of treatment were measured using the Radiation Treatment and Oncology Groups (RTOG) and LENT SOM systems. The results showed that freedom from PSA failure was higher in the 74 Gy group compared to the 64 Gy group, but this did not reach conventional levels of statistical significance with 5-year actuarial control rates of 71% (95% CI 58–81%) in the 74 Gy group vs 59% (95% CI 45–70%) in the 64 Gy group. There were 23 failures in the 74Gy group and 33 in the 64 Gy group (Hazard ratio 0.64, 95% CI 0.38–1.10, P=0.10). No difference in disease control was seen between the 1.0 and 1.5 cm margin groups (5-year actuarial control rates 67%, 95% CI 53–77% vs 63%, 95% CI 50–74%) with 28 events in each group (Hazard ratio 0.97, 95% CI 0.50–1.86, P=0.94). Acute side effects were generally mild and 18 weeks after treatment, only four and five of the 126 men had persistent ⩾Grade 1 bowel or bladder side effects, respectively. Statistically significant increases in acute bladder side effects were seen after treatment in the men receiving 74 Gy (P=0.006), and increases in both acute bowel side effects during treatment (P=0.05) and acute bladder sequelae (P=0.002) were recorded for men in the 1.5 cm margin group. While statistically significant, these differences were of short duration and of doubtful clinical importance. Late bowel side effects (RTOG⩾2) were seen more commonly in the 74 Gy and 1.5 cm margin groups (P=0.02 and P=0.05, respectively) in the first 2 years after randomisation. Similar results were found using the LENT SOM assessments. No significant differences in late bladder side effects were seen between the randomised groups using the RTOG scoring system. Using the LENT SOM instrument, a higher proportion of men treated in the 74 Gy group had Grade ⩾3 urinary frequency at 6 and 12 months. Compared to baseline scores, bladder symptoms improved after 6 months or more follow-up in all groups. Sexual function deteriorated after treatment with the number of men reporting some sexual dysfunction (Grade⩾1) increasing from 38% at baseline to 66% at 6 months and 1 year and 81% by year 5. However, no consistent differences were seen between the randomised groups. In conclusion, dose escalation from 64 to 74 Gy using conformal radiotherapy may improve long-term PSA control, but a treatment margin of 1.5 cm is unnecessary and is associated with increased acute bowel and bladder reactions and more late rectal side effects. Data from this randomised pilot study informed the Data Monitoring Committee of the Medical Research Council RT 01 Trial and the two studies will be combined in subsequent meta-analysis.

Parecoxib, Valdecoxib, and Cardiovascular Risk

Selective inhibitors of cyclooxygenase (COX)-21 depress prostacyclin (PGI2) but not COX-1–derived thromboxane A2. The effects of thromboxane A2 would be exaggerated during treatment with COX-2 inhibitors, potentially predisposing patients to heart attack and stroke.2 A randomized controlled trial in patients undergoing coronary artery bypass graft (CABG) surgery and receiving either placebo or 40 mg parecoxib (Dynastat; Pfizer), the intravenously administered prodrug of valdecoxib (Bextra; Pfizer), followed by oral valdecoxib 40 mg BID for 14 days, revealed a cluster of cardiovascular events.3 In a second study, one group received parecoxib/valdecoxib, another …

Cerebral palsy and socioeconomic status: a retrospective cohort study

Aims: To study the relation between risk of cerebral palsy and socioeconomic status. Methods: A total of 293 children with a diagnosis of cerebral palsy out of 105 760 live...

Linkage disequilibrium mapping of bipolar affective disorder at 12q23-q24 provides evidence for association at CUX2 and FLJ32356.

Chromosome 12q23-q24 has been implicated by several linkage studies as harboring a gene for bipolar affective disorder. We performed linkage disequilibrium (LD) mapping with 17 microsatellite markers across a 1.6 Mb-wide segment forming the central part of our narrowest linkage region. A significant signal (P = 0.0016) was identified for one microsatellite marker in our UK Caucasian case-control sample (347 cases, 374 controls). Genes, including regulatory elements, around this marker were screened for mutations and the LD structure of the region determined by genotyping 22 SNPs and insertion/deletion polymorphisms in 94 individuals. A set of 11 haplotype tagging (ht) SNPs was genotyped in our sample using a two-stage procedure. Two SNPs (rs3847953 and rs933399) and an insertion/deletion with putative functional relevance (which are in high LD with each other and with the microsatellite marker) showed significant or nearly significant association with bipolar disorder after Bonferroni-correction (reaching nominal P values from P = 0.002 to P = 0.005). In a sample of 110 UK Caucasian parent-offspring trios there was a trend for an over transmission in the same direction that failed to meet conventional levels of statistical significance. Our data provide evidence for association between bipolar mood disorder and markers on chromosome 12q23-q24 but need replication in independent samples.

Depot bromperidol decanoate for schizophrenia.

Antipsychotic drugs are the mainstay treatment for schizophrenia. Long-acting depot injections of drugs such as bromperidol decanoate are extensively used as a means of long-term maintenance treatment. To assess the effects of depot bromperidol versus placebo, oral antipsychotics and other depot antipsychotic preparations for people with schizophrenia in terms of clinical, social and economic outcomes. Relevant trials were identified by searching Biological Abstracts (1982-1999), Cochrane Library (Issue 2, 1999), Cochrane Schizophrenia Group's Register (May 1999), EMBASE (1980-1999), MEDLINE (1966-1999) and PsycLIT (1974-1999). References of all identified trials were inspected and Janssen-Cilag was contacted in order to identify more trials. An update search was undertaken in October 2003. The Schizophrenia Groups trials register is based on regular searches of BIOSIS Inside; CENTRAL; CINAHL; EMBASE; MEDLINE and PsycINFO; the hand searching of relevant journals and conference proceedings, and searches of several key grey literature sources. A full description is given in the Group's module. All randomised trials focusing on people with schizophrenia where depot bromperidol, oral antipsychotics or other depot preparations were sought. Primary outcomes were death, clinically significant change in global function, mental state, relapse, hospital admission, adverse effects and acceptability of treatment. Data were extracted independently by two reviewers and cross-checked. Fixed effects relative risks (RR) and 95% confidence intervals (CI) were calculated for dichotomous data. Weighted or standardised means were calculated for continuous data. Where possible, the number needed to treat statistic (NNT) was calculated. Analysis was by intention-to-treat. Four controlled clinical trials were included (total n=117). We identified a single small study of six months duration comparing bromperidol decanoate with placebo injection. Similar numbers left the study before completion (n=20, 1 RCT, RR 0.4 CI 0.1 to 1.6) and there was no clear differences between bromperidol decanoate and placebo for a list of adverse effects (n=20, 1 RCT, RR akathisia 2.0 CI 0.21 to 18.69, RR increased weight 3.0 CI 0.14 to 65.9, RR tremor 0.33 CI 0.04 to 2.69). When bromperidol decanoate was compared with fluphenazine depot we found no important change on global outcome (n=30, RR no clinical important improvement 1.50 CI 0.29 to 7.73). People allocated to fluphenazine decanoate and haloperidol decanoate had less relapses than those given bromperidol decanoate (n=77, RR 3.92 Cl 1.05 to 14.60, NNH 6 CI 2 to 341). People allocated bromperidol decanoate required additional antipsychotic medication somewhat more frequently than those taking fluphenazine decanoate and haloperidol decanoate but the results did not reach conventional levels of statistical significance (n=77, 2 RCTs, RR 1.72 CI 0.7 to 4.2). The use of benzodiazepine drugs was very similar in both groups (n=77, 2 RCTs, RR 1.08 CI 0.68 to 1.70). People left the bromperidol decanoate group with the same frequency as those allocated other depots (n=97, 3 RCTs, RR 1.92 CI 0.8 to 4.6). Anticholinergic adverse effects were equally common between bromperidol and other depots (n=47, RR 3.13 CI 0.7 to 14.0) and additional anticholinergic medication was needed with equal frequency in both depot groups, although results did tend to favour the bromperidol decanoate group (n=97, 3 RCTs, RR 0.80 CI 0.64 to 1.01). The incidence of movement disorders was similar in both depot groups (n=77, 2 RCTs, RR 0.74 CI 0.47 to 1.17). Currently, minimal poorly reported trial data suggests that bromperidol decanoate may be better than placebo injection but less valuable than fluphenazine or haloperidol decanoate. If bromperidol decanoate is available it may be a viable choice, especially when there are reasons not to use fluphenazine or haloperidol decanoate. Well-conducted and reported randomised trials are needed to inform practice in Belgium, Germany, Italy and the Netherlands.

Polychemotherapy (CT) in pre- and post-menopausal women with early breast cancer prescribed 5 years tamoxifen (T) –results from the UK NCRI Adjuvant Breast Cancer (ABC) international trial in 1991 patients

656 Background: EBCTCG 2000 systematic overview confirmed overall survival advantage of CT in early breast cancer (BC), but doubts remain over these benefits in presence of concomitant prolonged tamoxifen (T) and ovarian ablation (OA) in premenopausal women. Methods: The ABC trial is a pragmatic, randomised phase III trial. One aspect of ABC compared CT (any type but CMF recommended) vs. none in patients with early (1–3a) BC. All patients prescribed 20mg T for 5 years. Patients could receive OA or not as elective treatment. Main endpoints: relapse-free survival (RFS) and overall survival (OS). Results: During 1993 to 2000, 1991 (987 CT, 1004 no CT) patients were randomised from 106 UK and 16 non-UK centres. Pre-treatment patient characteristics were well-balanced. In total 62% patients were age >50 years, 56% node positive, and 64% oestrogen receptor (ER) positive. 92% patients received CT as allocated. Type of CT was CMF (87%), anthracycline containing (11%). 244 patients received OA. Results based on median follow-up of 5.0 years, 601 relapses and 485 deaths. CT suggested an improvement in RFS (HR 0.86 [95%CI 0.73–1.01] p=0.061) and OS (HR 0.87 [95%CI 0.73–1.04] p=0.12) but unadjusted results did not reach conventional levels of statistical significance. Adjusting for nodal status, ER, age, OA slightly increases magnitude of benefit of CT (RFS: HR 0.83 (0.71–0.97) p=0.024, OS: HR0.84 (0.70–1.00) p=0.051). Results (RFS) were independent of ER status. Impact appeared greater for women aged <50 (<50 HR 0.76 [0.57–1.01], ≥50 HR 0.89 [0.73–1.09]), for those (pre-menopausal) not receiving OA (without OA HR 0.71 [0.47–1.08], with OA HR 0.94 [0.56–1.56]) and for those from centres with good CT compliance (good CT HR 0.83 [0.68–1.01], poor CT HR 0.95 [0.71–1.25]). Conclusion: This is the first report of CT in women with early BC receiving concomitant and 5-year duration T, showing that benefits of CT are retained in this context. No significant financial relationships to disclose.

Polarity therapy and cancer-related fatigue

8147 Background: Cancer and its treatment produces fatigue in a majority of patients. Despite the very high prevalence of this problem, no effective treatments are available. Increasing numbers of cancer patients are turning to complementary and alternative medicine practices to find relief. We examined the efficacy of Polarity Therapy (PT), a massage-like intervention classified by NCCAM as an “energy therapy,” in relieving fatigue in female breast cancer patients receiving radiation therapy. Methods: 15 Patients (mean age = 52.5) were randomly assigned to one of three trial groups. Arm 1 patients did not receive any (PT). Patients assigned to Arm 2 received a PT tx on a Tuesday during their 4th week of radiation treatment. Patients assigned to Arm 3 received PT on Tuesdays during \(\underline{\mathrm{both}}\) their 4th and 5th weeks of radiation treatment. Patients completed the Brief Fatigue Inventory (BFI) during the 3rd through 5th weeks of radiation treatment. Each PT session lasted 60–75 minutes with the patient lying fully clothed on a massage table while the same experienced, licensed Polarity Therapist administered all PT treatments. Results: Change scores on the BFI from baseline showed a statistically significant improvement in fatigue in the 10 patients who had PT compared to the 5 patients who did not after the first week (mean change score: one PT tx = 1.0, SD = 1.46; no PT = −0.6, SD = .93; p = .05). Eight of the ten patients showed an improvement in fatigue following their first PT treatment, while only one of five patients in the control group had an improvement in fatigue during the same time frame. A one-way analysis of variance (ANOVA) examining group differences the following week closely approached conventional levels of statistical significance (mean change score: two PT tx = 2.2, SD = 1.03; one PT tx = 1.6, SD = 1.10; no PT = 0.6, SD = .87; p = .08. A significant (p = .03) linear contrast suggested a possible dose effect. Conclusions: This pilot supports a confirmatory trial of Polarity Therapy for cancer- related fatigue. No significant financial relationships to disclose.

Information cascades and the distribution of economic recessions in capitalist economies

We consider in this paper the distribution of the cumulative size of recessions in 17 capitalist countries over the period 1871–1994, using data on annual percentage changes in real GDP. A recession is defined as a year in which GDP growth is negative, and the cumulative change is the change from peak to trough during a recession period. We examine both the whole sample and different partitions of the data. The null hypothesis that the size distribution of recessions follows an exponential distribution is never rejected at the conventional level of statistical significance, p=0.05. However, there are always a small number of large recessions, no matter how the data is partitioned, which are not well fitted by a least-squares regression of the log of size on the rank of size. In other words, in a qualitative sense we see a bimodal distribution of recessions, with an exponential fit to the bulk of the data, and a second peak describing a small number of very large recessions. A previously published agent-based economic theory model of the business cycle, calibrated purely on US data, is able with no change in its parameters to generate an exponential distribution of the size of recessions very close to that which is actually observed. In this model, information flows between agents on a completely connected network.

Is the season of birth association with psychosis due to seasonal variations in foetal growth or other related exposures? A cohort study.

To investigate the association between season of birth and psychosis, and to assess whether any association is caused by seasonal fluctuations in foetal growth or other related exposures. Cohort of 747 432 Swedish males and females born between 1973 and 1980 and followed up from 16 years of age to 31 December 1999. Psychiatric admissions were identified using the Swedish Inpatient Discharge Register. The analysis is based on the 696 025 subjects with complete data. A total of 506 (0.07%) subjects developed schizophrenia and 879 (0.13%) non-affective non-schizophrenic psychoses. There was a moderate increased risk of schizophrenia amongst winter births, hazard ratio 1.23 (95% confidence interval 0.96-1.59), but this did not reach conventional levels of statistical significance. There was no association with non-affective psychoses. We found no evidence that associations were confounded by measures of foetal growth or maternal socioeconomic position. There was no evidence that seasonal effects on schizophrenia differed in men and women. Season of birth associations with schizophrenia do not appear to be confounded by birth-related exposures.

Religion, Culture, and Economic Performance

The hypothesis that the coefficients on variables of religious affiliation are jointly equal to zero can frequently be rejected at conventional levels of statistical significance (i.e., religion matters), but no robust relationship between adherence to major world religions and national economic performance is uncovered, using both cross-national and subnational data. The results with respect to Islam do not support the notion that it is inimical to growth. On the contrary, virtually every statistically significant coefficient on Muslim population shares reported in this paper—in both cross-country and within-country statistical analyses—is positive. If anything, Islam promotes growth.

Life course influences on quality of life in early old age

A growing literature demonstrates life course influences on health in early old age. The present paper is the first to examine whether similar processes also influence quality of life in early old age. The question is theorised in terms of structured dependency and third age, and the life course pathways by which people arrive at these destinations in later life. The issues are investigated in a unique data set that contains health and life course information on some 300 individuals mostly aged 65–75 years, enhanced in 2000 by postal survey data on quality of life. Several types of life course effect are identified at conventional levels of statistical significance. Long-term influences on quality of life, however, are less marked than those on health. Quality of life in early old age appears to be influenced primarily by current contextual factors such as material circumstances and serious health problems, with the influence of the life course limited mostly to its shaping of an individual's circumstances in later life. The implication for policy is that disadvantage during childhood and adulthood does not preclude good quality of life in early old age.

User's guide to the Chiropractic Literature-IB: how to use an article about therapy

User's guide to the Chiropractic Literature-IB: how to use an article about therapy

Height, Leg Length, and Cancer: The Caerphilly Study

The incidence of several cancers increases with height. Some studies report that leg length, a marker of prepubertal growth, is the component of height underlying these associations, but few prospective studies have investigated this issue. We examined height-cancer associations in a 21-yr follow-up of a cohort based on 2,512 men aged 45-59 living in the town of Caerphilly, South Wales, U.K., between 1979 and 1983. The men underwent a detailed examination, and 2,393 (95.3%) had measures of height and sitting height (from which leg length and trunk length were derived) recorded and were cancer-free at baseline. Cox proportional hazards models were used to investigate associations of height, leg length, and trunk length with overall and site-specific cancer incidence; 328 men developed cancer over the follow-up period. Associations with height were weak and did not reach conventional levels of statistical significance. In models controlling for age, socioeconomic position, smoking, and body mass index, cancer incidence was weakly positively associated with increases in height; hazard ratio (HR, 95% confidence interval) per standard deviation (6 cm) increase in height was 1.09 (0.97, 1.21; P = 0.14). Associations were somewhat stronger in relation to leg length (HR 1.09; (0.97-1.22) than trunk length (HR 1.05; 0.94-1.18). Height-cancer associations were strongest for lung cancer (HR 1.21; 0.96-1.51). This analysis provides limited support for the suggestion that leg length is the component of height underlying height-cancer associations.

Religion, Culture and Economic Performance

The hypothesis that the coefficients on variables of religious affiliation are jointly equal to zero can frequently be rejected at conventional levels of statistical significance (i.e., religion matters), but no robust relationship between adherence to major world religions and national economic performance is uncovered, using both cross-national and subnational data. The results with respect to Islam do not support the notion that it is inimical to growth. On the contrary, every statistically significant coefficient on Muslim population shares reported in this paper - in both cross-country and within-country statistical analyses - is positive. If anything, Islam promotes growth.

Major psychiatric disorders and the serotonin transporter gene (SLC6A4): family-based association studies.

The serotonin transporter (5-HTT) is a suitable candidate gene to test for involvement in the pathogenesis of major psychiatric disorders. We used the method of family-based controls to test for association between disease and a variable number tandem repeat (VNTR) in intron 2 of the gene, which has received support for involvement in the pathogenesis of several psychiatric disorders. We analysed 413 proband-parent trios of Bulgarian origin: 266 had a schizophrenic proband, 103 had a bipolar proband and 44 had a schizoaffective proband. The results were analysed using the extended transmission disequilibrium test. Possible effects of different alleles on certain clinical variables were examined by correlation analysis. Three alleles were detected: STin2.9, STin2.10 and STin2.12. None of the three diagnostic samples showed preferential transmission of alleles that reached conventional levels of statistical significance. We could not confirm previous results that STin2.12 allele increases susceptibility to bipolar disorder type I. The rare STin2.9 showed a non-significant trend for preferential transmission in the sample as a whole: 18 transmitted versus 11 non-transmitted (P = 0.2). The VNTR polymorphism in the 5-HTT gene does not appear to be a major risk factor for increasing susceptibility to major psychiatric disorders.

Stereotype Threat Effects on the Raven Advanced Progressive Matrices Scores of African Americans1

The present experiment examined stereotype threat effects on the Raven Advanced Progressive Matrices (Raven) scores of African Americans. Support was found for Hypothesis 1, which stated that African Americans would experience significantly greater stereotype threat than Whites during an IQ testing situation. Hypothesis 2 proposed that a significant Race x Test Diagnosticity Condition interaction would occur in which the mean difference in intelligence test scores between African Americans and Whites (favoring the latter group) would be largest when the test was framed as a measure of intelligence, and smaller when framed as nonindicative of intelligence. Limited support for Hypothesis 2 was found in that a marginally significant Race x Test Diagnosticity Condition interaction was obtained. However, the interaction failed to reach conventional levels of statistical significance. Although preliminary, the present findings provide some support for the notion that stereotype threat compromises the intelligence test performance of African Americans.

Evaluating Social Policy by Experimental and Nonexperimental Methods

Although it is important to establish causal relationships in social policy evaluation, the effects are difficult to observe due to sample selection. To evaluate the performance of estimators designed to handle sample selection bias, we analyse data from a Norwegian rehabilitation project with a randomised experimental design. The data permit us to compare the performance of different nonexperimental estimators with the experimental results. In our case study we find that nonexperimental evaluation based on sample selection estimators with selection terms that fail to meet conventional levels of statistical significance is highly unreliable. The difference in difference estimator and propensity score matching estimators perform better in our context. JEL classification: C51; J24; H43; I12

Editor's Choice

Editor's Choice