Statistical Proof Research Articles

Fashioning the 1950s “Vassar Girl”: Vassar Student Identity and Campus Dress, 1947–60

In 1957, Clifford Coffin photographed real Vassar students as models for the annual Vogue college issue. High fashion and elite education were combined and the photographs underscore the significance of the “Vassar Girl” as both a cultural and sartorial icon during this decade: a media-cultivated, almost mythical presence that permeates 1950s American media, popular culture, and fashion discourses. What these images more pressingly demonstrate is that, by the 1950s, the “Vassar Girl” was a powerful entity in American culture: seemingly both real and imagined, but lucrative all the same. Genuine Vassar students had to navigate a series of sartorial assumptions and expectations to assert and formulate their own identity, both within and without, the campus culture, providing material for a rich exploration of individual and collective identity of Vassar students at the mid-century.Following the collection and analysis of original oral testimony of hundreds of mid-century Vassar graduates and studies of iconic garments of the period, this article provides an explorative analysis of the ways in which fashion functioned in the construction of personal and collective identity at Vassar, and the construction and perpetuation of a “Vassar Girl” aesthetic and brand in American fashion and media discourses of the 1950s. Grounding the exploration in an analysis of Vassar's revolutionary psychological and developmental study of its students during the 1950s (the Mellon Study), this article will undercut media discourses and hyperbole to root considerations of dress and identity in the statistical and narrative proof of original college records from the pre-feminist era, which was also the last full decade of all-female education at Vassar.

The impact of wine franchise laws on consumer choice and pricing

PurposeThe purpose of this paper is to compare and contrast wine sales in neighboring franchise law and non‐franchise law states in order to determine impact on wine price, consumer choice, consumer satisfaction, and stakeholder perception.Design/methodology/approachThe study used qualitative interviews with 14 wineries, distributors, and retailers, statistical analysis of Nielsen Scantrack data, and an online survey of 401 wine consumers in Georgia and Florida, USA.FindingsResults show statistical proof that Florida offers more wine selection and lower wine prices on matching brands than Georgia. Qualitative interviews indicate wineries, distributors, and retailers perceive differences in wine choice, price, and overall operating costs in these two states. However, there was no statistical difference between a sample of 401 consumers from Georgia and Florida when asked about their satisfaction level with wine choice and pricing within their state.Research limitations/implicationsFor practical purposes, the research was limited to only two US states. It would be useful to duplicate this study in other states.Practical implicationsPractical implications include the need for new wineries desiring to enter franchise law states to carefully research regulations and distributors before making a commitment, as well as the social issue of less wine choice and higher prices for consumers in Georgia versus Florida.Originality/valueThis is the first empirical study in the USA to focus on the impact of wine franchise laws on consumer choice and wine price. It yields useful information that contributes to the body of knowledge for wine and policy research.

The Riemannian Structure of the Three-Parameter Gamma Distribution

In this paper, we will utilize the results already known in differential geometry and provide an intuitive understanding of the Gamma Distribution. This approach leads to the definition of new concepts to provide new results of statistical importance. These new results could explain Chen [1-3] experienced difficulty when he attempts to simulate the sampling distribution and power function of Cox’s [4,5] test statistics of separate families of hypotheses. It may also help simplify and clarify some known statistical proofs or results. These results may be of particular interest to mathematical physicists. In general, it has been shown that the parameter space is not of constant curvature. In addition, we calculated some invariant quantities, such as Sectional curvature, Ricci curvature, mean curvature and scalar curvature.

Investigations on non‐inferiority—the Food and Drug Administration draft guidance on treatments for nosocomial pneumonia as a case for exact tests for binomial proportions

This paper addresses statistical issues in non-inferiority trials where the primary outcome is a fatal event. The investigations are inspired by a recent Food and Drug Administration (FDA) draft guideline on treatments for nosocomial pneumonia. The non-inferiority margin suggested in this guideline for the endpoint all-cause mortality is defined on different distance measures (rate difference and odds ratio) and is discontinuous. Furthermore, the margin enables considerable power for the statistical proof of non-inferiority at alternatives that might be regarded as clinically unacceptable, that is, even if the experimental treatment is harmful as compared with the control. We investigated the appropriateness of the proposed non-inferiority margin as well as the performance of possible test statistics to be used for the analysis. A continuous variant of the margin proposed in the FDA guideline together with the unconditional exact test according to Barnard showed favorable characteristics with respect to type I error rate control and power. To prevent harmful new treatments from being declared as non-inferior, we propose to add a 'second hurdle'. We discuss examples and explore power characteristics when requiring both statistical significance and overcoming the second hurdle.

Early versus Delayed Surgical Treatment in Open Hand Injuries: A Paradigm Revisited

To date it remains unclear if a delayed surgical treatment of open hand injuries after more than 6h may be detrimental to outcome. Previous investigations by McLain et al. (J Hand Surg Am 16:108-112, 1980 9), Nylen and Carlsson (Scand J Plast Reconstr Surg 14:185-189, 1991 10) could not find statistical proof of correlation between infection rate and delayed surgical treatment after open hand injuries up to 18h. The current study was designed to investigate the outcome of early versus delayed surgical treatment after open hand injury. A retrospective data analysis of all patients sustaining an open hand injury between January 1 and December 31 2006 was performed. Patients with incomplete data records were excluded. Patients were stratified according to time delay of surgical treatment and injury complexity. Complications, revision rate, and functional outcome were also investigated. Between January 1 and December 31 2006 a total of 458 patients with open hand injuries were treated at the University Hospital of Berne, Switzerland. The records were retrospectively analyzed and a subgroup of 100 patients were randomly determined for standardized follow-up evaluation. There were no significant differences regarding infection, complication, and revision rate between early (<6h) and delayed (6-24h) surgical treatment of open hand injuries. Independently from the time point of surgical treatment, patients with complex injuries and longer operative times had an increased infection rate (p=0.05) and revision rate (p=0.003). Delayed surgical treatment (6-24h) of open hand injuries did not increase infection or revision rates in open non-devascularizing hand injuries.

Monitoring mortality trends in low-resource settings

Background Recent calls for continuous monitoring of large-scale public health interventions in low-income countries (1) have coincided with the increased use of routine monitoring for outcomes after episodes of care in high-income settings. (2-5) Graphical methods have become available to monitor short-term outcomes adjusted for patients' risk. (2,3) However, there are no documented examples of these tools being used to monitor mortality outcomes in low-income settings. As well as facilitating routine monitoring activities, graphical displays of outcomes with time can also suggest avenues for academic research by enhancing the interpretation of data. (1) This in turn can shed light on the potential mechanisms that underlie the observed effects. Here we discuss the use in a low-income setting of one such graphical method, the variable life adjusted display (VLAD). Its advantages are that the charts are easy to understand and can alert users quickly if the outcomes observed are worse than predicted. We briefly describe this method and its uses, and illustrate how it was used with data from a recent trial that studied neonatal deaths in India. (6) Use of VLAD The VLAD method (4) was devised to present time-ordered peri-operative outcomes after cardiac surgery, adjusted for individual patients' risk and based on an established risk scoring system. It has since been adopted internationally as a standard monitoring tool for mortality after cardiac surgery and in many other contexts. (5) The VLAD method does not provide statistical proof of an effect; however, it does provide an excellent mechanism for rapidly identifying unusually favourable or unfavourable outcomes that might warrant study. VLAD method Here we provide a brief description of the technique. A full description is available at http://www.ucl.ac.uk/ operational-research/Analysis Tools/ VLAD and elsewhere. (4,5) The method is suitable for any binary short-term outcome; here we consider neonatal survival after live birth. A VLAD chart shows the difference between the expected numbers of deaths and the numbers of deaths observed with time. As originally designed, the expected numbers of deaths with time was determined by estimating individual risk, but the same risk of death can be used for all individuals. This latter approach is well suited to monitoring individual-level outcomes after public health interventions. For these contexts risk of death can be an observed or target mortality rate. If the baseline probability of neonatal death is given by P, where P is a number between zero and 1 (e.g. a mortality rate of 20% would yield P = 0.2), then on average every live birth would be expected to result in P neonatal deaths. By assigning a value of 1 to an observed neonatal death and a value of zero to an observed survival, the score associated with each live birth can be calculated as (expected outcome) minus (observed outcome). Thus, if a baby dies, that birth is associated with a score of P - 1 (less than zero), and if a baby survives then that birth is associated with a score of P (greater than zero). The VLAD score is the cumulative total of scores with time and represents the cumulative difference between expected and observed deaths. If this difference is greater than zero there have been fewer deaths than expected, and if it is less than zero then there have been more deaths than expected. A VLAD chart is constructed by plotting the VLAD score with time, which reveals trends in outcomes. Subgroups (e.g. trial arms, different geographical regions or individual hospitals) can be plotted separately, with multiple traces on the same VLAD chart. We illustrate the use of VLAD with data from an intervention that involved local women's groups discussing issues related to maternal and newborn health in eastern India. The intervention was rigorously evaluated in a cluster-randomized controlled trial which ran from 31 July 2005 to 30 July 2008, and resulted in a 32% (odds ratio: 0. …

A software tool for monitoring legal minimum length of landings: Case study of a fishery in southern Spain

Abstract The regulation of minimum legal size (MLS) of catches is a tool widely applied in the management of fisheries resources, although the MLS does not always coincide with the length at first maturity (LFM). The optimization of this management tool requires a series of quality control in fish markets and transportation. A software application has been developed to make the control of the landings of several target species easier and faster. In order to test and make this tool operational, six species of commercial interest were selected: four species of fish and two species of bivalves. It is proposed to estimate the proportion of illegal specimens in the studied lot from the proportion of illegal individuals found in the samples taken from this lot. The input data for the application are the minimum legal size (MLS) of the species and the total length (TL) of each specimen sampled. The output data is a statistical summary of the percentage of specimens of size less than the legal minimum (TL≤MLS) within different confidence intervals (90%, 95% and 99%). The software developed will serve as a fast, efficient and easy to manage tool that allows inspectors to determine the degree of compliance on MLS control and to make a decision supported by statistical proof on fishing goods.

Did Technology Shocks Drive the Great Depression? Explaining Cyclical Productivity Movements in U.S. Manufacturing, 1919–1939

Technology shocks and declining productivity have been advanced as important factors driving the Great Depression in the United States, based on real business cycle theory. We estimate an improved measure of technology for interwar manufacturing, using data from the U.S. census reports. There is clear evidence of increasing returns to scale and we find no statistical proof that technology shocks led to changes in hours worked or other inputs. This contradicts a key prediction of real business cycle theory. We find that increasing returns to scale are not due to market power but to labor and capital hoarding.

The fieldwork model: An anthropological perspective on the process of change in long-term psychoanalytic psychotherapy

The question of what takes place in the psychoanalytic consulting room, and whether it has any therapeutic effect on the patient, has consistently been behind various criticisms of the long-term psychoanalytically informed therapies. When it comes to goal-oriented talking therapies –the only ones favoured by the state mechanism – statistical proof of their efficacy is thought of as proof of their scientific value. However, disciplines such as social anthropology have consistently based their complex understanding of social phenomena in qualitative data. This article draws upon the anthropological fieldwork model. It also looks at how anthropological theory has evolved as a result of the increasing centrality of the fieldwork model and a gradual acceptance of the inter-subjective process. It aims to highlight how lived experience, the paradigm on which the fieldwork model is based, constitutes a highly effective way of understanding in depth the patient and the patient's history. Such understanding is in turn the main factor behind profound and lasting psychological change. It is argued here that psychotherapy research and practice could be enhanced by following the example of anthropological qualitative research.

Significance analysis and statistical dissection of variably methylated regions

It has recently been proposed that variation in DNA methylation at specific genomic locations may play an important role in the development of complex diseases such as cancer. Here, we develop 1- and 2-group multiple testing procedures for identifying and quantifying regions of DNA methylation variability. Our method is the first genome-wide statistical significance calculation for increased or differential variability, as opposed to the traditional approach of testing for mean changes. We apply these procedures to genome-wide methylation data obtained from biological and technical replicates and provide the first statistical proof that variably methylated regions exist and are due to interindividual variation. We also show that differentially variable regions in colon tumor and normal tissue show enrichment of genes regulating gene expression, cell morphogenesis, and development, supporting a biological role for DNA methylation variability in cancer.

Balancing Access and Evaluation in the Approval of New Cancer Drugs

PATIENTS WITH CANCER WHO LACK EFFECTIVE TREATment desperately seek novel interventions that might extend their lives. They do not have time to wait for the US Food and Drug Administration’s (FDA’s) regular approval process, which requires that sponsors demonstrate “substantial evidence of clinical benefit (or efficacy) from adequate and well-controlled investigations.” Clinicians, for their part, are eager to offer “rational” treatments targeted at the molecular signatures of disease to their sickest patients. Concern about the plight of desperate patients, along with the promise of targeted treatments, makes it easy to lose sight of the public interest in rigorous evaluation of experimental drugs before they are marketed to patients. Understandable focus on patients with serious illness diverts attention from the value to present and future patients and to society of ensuring the efficacy and safety of new treatments. Moreover, the case for rigorous evaluation rests on technical issues of study end points and statistical proof that are not widely understood or appreciated. How should the interests of individual patients and clinicians in access to novel cancer treatments be weighed against the public interest in rigorous evaluation? There is no easy solution to this dilemma. Any sound public policy must balance 2 types of error: speeding access to treatments that are not beneficial or have an unfavorable benefitharm ratio and unreasonably delaying access to beneficial treatments. The second error receives the most public attention. The first, however, is also critically important for current and future patients. In addition, society has a stake in the rigorous evaluation of new cancer drugs, especially in view of escalating costs. A major regulatory initiative introduced in 1992 in response to perceptions that the time required for regular approval was excessive allows “accelerated approval” of new drugs that are judged likely to provide benefits, compared with available treatments, for life-threatening conditions. Approval may be granted following positive results of trials assessing surrogate end points such as tumor response or progression-free survival. Sponsors seeking accelerated approval must conduct follow-up randomized controlled trials (RCTs) evaluating clinical benefit (eg, survival) or validated surrogate outcomes for clinical benefit to obtain regular approval and continue marketing. How well does this mechanism resolve the access/evaluation dilemma? A recent FDA review found that between late 1992 and mid-2010, 35 oncology drugs were granted accelerated approval for 47 new indications. Evaluating the success of the accelerated approval program is challenging. Although the advantages of speeding access to drugs for which the clinical benefits are later confirmed are clear, the drawbacks are less visible. Accelerated approval is usually based on results from small, often single-group trials. Of the 47 accelerated approvals, 28 were based on single-group trials and 35 on response rate end points. Although regular approval occasionally is warranted on the basis of high response rates compared with historical controls, RCTs assessing survival or validated surrogates for survival remain the default standard for rigorous evaluation. In addition, early access in clinical practice inevitably makes it difficult to complete confirmatory trials in a timely fashion, as once an agent is clinically available to treat a lifethreatening disease, the conduct of an RCT comparing the agent to a control becomes both ethically and practically challenging. Historical examples from both within and outside oncology provide compelling evidence that restricting access to experimental treatments to well-designed trials is the most efficient way to develop the rigorous data needed to evaluate efficacy and safety. The success of the accelerated approval depends on both the extent to which follow-up RCTs reliably confirm benefit and the timely completion of these trials. The FDA review indicated that only 3 of 29 oncology drugs for which confirmatory trials were completed failed to demonstrate benefit. Subsequently, the FDA determined that the benefits of a fourth drug, which received accelerated approval in 2008 for a subset of patients with metastatic breast cancer, could not be confirmed. Furthermore, the FDA faces important disincentives to withdrawing approval based on either unfavorable evidence from confirmatory trials or failure to com-

Comments on “Plotting Positions in Extreme Value Analysis”

AbstractThis comment addresses the role of sampling error in extreme value analysis. A note published in this journal claimed that Weibull’s 1939 estimator for sample probability has a unique status that invalidates all other estimators and renders invalid all of the developments of unbiased distribution-dependent estimators made since 1939. The note concluded that the use of distribution-dependent estimators should be abandoned and that many estimates of the weather-related risks should be reevaluated and the related building codes and other related regulations updated. This comment uses rigorous statistical proofs to make the diametrically opposite case: namely, that development of distribution-dependent estimators has resulted in an improvement in accuracy over the past half century and that no changes are required to the basis of weather-related building codes and regulations. These rigorous proofs are supplemented by sampling experiments that demonstrate their validity. This comment provides an introduction to the basic statistical concepts of the statistical modeling of extremes, including unbiased estimators for the model parameters.

The ICF and Florence Nightingale - bringing data to statistical proof

The ICF and Florence Nightingale - bringing data to statistical proof

Statistical Methods Are Inadequate

I would agree that, when administering pharmacologically effective medications, the mechanisms that underlie the administration of a placebo should be consciously exploited. I would, however, argue that such mechanisms cannot be achieved with very little effort. To influence positively the conditional reflexes and expectations of a patient is not possible without a greater effort being involved—much in the same way as a suitable personal charisma on the doctor’s part and of the atmosphere where the treatment is being given. Doctors are not trained for this, and neither does the medical system expect this of them in everyday clinical practice. The placebo effect is still being confused with the phenomenon of prescribing a “sham medication.” However, placebo effects as I would describe them are so much more. No surgery, no irradiation treatment, no drug administration, no patient between doctor and patient, no expectation on the doctor’s part or the patient’s part are ever devoid of placebo or nocebo effects, even if this has not been (and probably cannot be) statistically proved in the individual case. These phenomena touch on sociological, philosophical, or even spiritual questions that statistical methods famously cannot capture. Doctors are prone to underestimating the placebo effect and often totally ignore the phenomenon, except when it is essential in drug experiments. Guideline based medicine, which is increasingly based on statistical proof of efficacy, is an obstacle to the access to and inclusion of placebo effects in everyday medical practice.

Ultrasound-Guided Regional Anesthesia and Patient Safety

The role of ultrasound-guided regional anesthesia (UGRA) in reducing the frequency of regional anesthetic-related complications is difficult to ascertain from analyzing the limited literature on the topic. This evidence-based review critically evaluates the contributions of UGRA to improved patient safety, particularly as compared with standard nerve localization tools. Randomized controlled trials that compared UGRA with another form of neural localization and case series of more than 500 patients were used to compare safety parameters. The quality of studies and strength of evidence were graded. Of those randomized controlled trials identified by our search techniques, 22 compared the incidence of postoperative nerve symptoms, 17 assessed local anesthetic systemic toxicity parameters, and 3 studied hemidiaphragmatic paresis. Statistical proof for meaningful reduction in the frequency of extremely rare complications, such as permanent peripheral nerve injury, is likely unattainable. Although there is evidence for UGRA reducing the occurrence of vascular puncture and the frequency of hemidiaphragmatic paresis, as yet there is at best inconclusive scientific proof that these surrogate outcomes are linked to actual reduction of their associated complications, such as local anesthetic systemic toxicity or predictable diaphragmatic impairment in at-risk individuals. This evidence-based review thus strives to summarize both the power and the limitations of UGRA as a tool for improving patient safety.

Impact of a diabetic foot care education program on lower limb amputation rate

BackgroundDiabetic foot complications are a leading cause of lower extremity amputation. With the increasing incidence of diabetes mellitus in the Arab world, specifically in the Kingdom of Saudi Arabia, the rate of amputation will rise significantly. A diabetic foot care program was implemented at King Abdulaziz Medical City in Riyadh, Saudi Arabia, in 2002. The program was directed at health care staff and patients to increase their awareness about diabetic foot care and prevention of complications. The purpose of this study was to perform a primary evaluation of the program’s impact on the rate of lower extremity amputation due to diabetic foot complications.MethodThis pilot study was the first analysis of the diabetic foot care program and examined two groups of participants for comparison, ie, a “before” group having had diabetic foot ulcers managed between 1983, when the hospital was first established, and 2002 when the program began and an “after group” having had foot ulcers managed between 2002 and 2004, in the program’s initial phase. A total of 41 charts were randomly chosen retrospectively. A data sheet containing age, gender, medical data, and the presentation, management, and outcome of diabetic foot cases was used for the analysis.ResultsThe before group contained 20 patients (17 males) and the after group contained 21 patients (16 males). There was no difference between the two groups with regard to age and comorbidities. The rate of amputation was 70% in the before group and 61.9% in the after group. There was a decrease in the percentage of toe amputation in the after group and an increase in the percentage of below-knee amputation in the before group. However, these changes were not significant.ConclusionThe program, although evaluated at an early stage, has increased the awareness of both patients and health care staff about the prevention and management of diabetic foot disease, and decreased the rate of lower extremity amputation. We believe that the statistical proof of its impact will be evident in the final evaluation.

Humoral and Cellular Immune Responses after Influenza Vaccination in Kidney Transplant Recipients

It has been speculated that influenza vaccination of renal allograft recipients could be associated with de novo production and/or increased titers of anti-HLA antibodies (HLA-Ab). To directly address this issue, we recruited 66 stable renal transplant recipients and 19 healthy volunteers during the 2005-2006 vaccination campaign. At day 0 and day 30 following vaccination, HLA-Ab were screened and in parallel influenza-specific antibody and T-cell responses were assessed. Humoral postvaccinal responses to A/H1N1 and A/H3N2 strains, but not B strain, were less frequent in transplanted patients than in control subjects. Significant expansion of influenza-specific IFN-gamma-producing T cells was observed at similar frequencies in patients and controls. There was no correlation between cellular and humoral postvaccinal responses. No impact of sex, age or immunosuppressive regimen could be evidenced. Vaccination was not associated with any significant change in preexisting or de novo anti-HLA sensitization. No episode of allograft rejection was recorded in any of the patients. Our results suggest that flu vaccination is safe in stable renal transplanted patients. Larger studies are needed for definitive statistical proof of the safety and effectiveness, with regard to the quality of the immune response, of yearly influenza vaccination in immunosuppressed patients.

Zero-Knowledge against Quantum Attacks

This paper proves that several interactive proof systems are zero-knowledge against general quantum attacks. This includes the well-known Goldreich-Micali-Wigderson classical zero-knowledge protocols for graph isomorphism and graph 3-coloring (assuming the existence of quantum computationally concealing commitment schemes in the second case). Also included is a quantum interactive proof system for a complete problem for the complexity class of problems having honest verifier quantum statistical zero-knowledge proofs, which therefore establishes that honest verifier and general quantum statistical zero-knowledge are equal: $\mathrm{QSZK}= \mathrm{QSZK}_{\mathrm{HV}}$. Previously no nontrivial interactive proof systems were known to be zero-knowledge against quantum attacks, except in restricted settings such as the honest verifier and common reference string models. This paper therefore establishes for the first time that true zero-knowledge is indeed possible in the presence of quantum information and computation.

History of the Peoples of the Great Hungarian Plain in the First Millennium: A Craniometric Point of View

We carried out an examination relying on six dimensions of 1,573 crania coming from the Great Hungarian Plain. The crania represent seven archeological periods: Sarmatian age (1-4th century), the period of transition (about 400-420), Hun and Gepidic epochs (about 420-455 and 455-567, respectively), early Avar age (about 568-670), late Avar period (about 670-895), the epoch of the Hungarian conquest and settlement (about 895-1000), and the Arpadian age (about 1000-1301). We were curious about the anatomical background behind cultural changes of the various populations that inhabited this area. After having noticed some discontinuities between the populations, as revealed by univariate analysis of single dimensions, we performed a principal-components analysis to see whether or not the diverse components showed eventual breaks in the sequence of the populations. Knowing that all the dominant populations had Asian roots, except for the Gepids of Germanic origin, we expected a considerable difference between the Gepidic population and all the other inhabitants. We also assumed that a conquest itself with a large-scale assimilation was unlikely to leave breaklike traits in anatomical patterns, except for aggressive conquests. We found that the second principal component (which correlated with cranial breadth and partly with height) showed a remarkable hiatus in both sexes between Gepids and early Avars. Having done a statistical proof (simultaneous tests for general linear hypotheses) of the observed phenomenon, we found that the gap referring to subsequent populations was significant only in males. A possible reason for this result is that the Avar conquest was much more radical than has been thought. In addition, considering that men were more likely to die in wars, women survived and were assimilated into the conquerors' populations with higher probability, so it is not surprising that the results of multicomparison tests are significant only in men.

History of the Peoples of the Great Hungarian Plain in the First Millennium: A Craniometric Point of View

We carried out an examination relying on six dimensions of 1,573 crania coming from the Great Hungarian Plain. The crania represent seven archeological periods: Sarmatian age (1–4th century), the period of transition (about 400–420), Hun and Gepidic epochs (about 420–455 and 455–567, respectively), early Avar age (about 568–670), late Avar period (about 670–895), the epoch of the Hungarian conquest and settlement (about 895–1000), and the Arpadian age (about 1000–1301). We were curious about the anatomical background behind cultural changes of the various populations that inhabited this area. After having noticed some discontinuities between the populations, as revealed by univariate analysis of single dimensions, we performed a principal-components analysis to see whether or not the diverse components showed eventual breaks in the sequence of the populations. Knowing that all the dominant populations had Asian roots, except for the Gepids of Germanic origin, we expected a considerable difference between the Gepidic population and all the other inhabitants. We also assumed that a conquest itself with a large-scale assimilation was unlikely to leave breaklike traits in anatomical patterns, except for aggressive conquests. We found that the second principal component (which correlated with cranial breadth and partly with height) showed a remarkable hiatus in both sexes between Gepids and early Avars. Having done a statistical proof (simultaneous tests for general linear hypotheses) of the observed phenomenon, we found that the gap referring to subsequent populations was significant only in males. A possible reason for this result is that the Avar conquest was much more radical than has been thought. In addition, considering that men were more likely to die in wars, women survived and were assimilated into the conquerors’ populations with higher probability, so it is not surprising that the results of multicomparison tests are significant only in men.