On Edwards' criterion of seasonality and a non-parametric alternative.

Hewitt, David; Milner, J; Csima, Adele; Pakula, A

doi:10.1136/jech.25.3.174

Cited by 69 publications

(74 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Seasonality would also be indicated on Edwards' (1961) test or on the test of Hewitt et al (1971) if there were an increasing (or a decreasing) secular trend in the numbers of cases registered per month: if incidence was increasing, the number of cases occurring per month early in a year would tend on average to be lower than in the later months of the same year (or, if incidence was decreasing there would tend to be more cases in early months than in late months of a year). Hinds et al (1981) noted that the average rate of increase in melanoma diagnoses in Hawaii during their study would have given a relatively small difference between JanuaryFebruary and November-December numbers of cases, and they did not correct for it in their analysis.…”

Section: Discussionmentioning

confidence: 99%

“…The data sources and methods used by the registry have been detailed by Hunt (1976 judged from various indirect measures, appear generally to be high (Doll et al, 1970;Waterhouse et al, 1976Waterhouse et al, , 1982Scott, 1983 Harmonic seasonality of first attendance was tested by the method of Edwards (1961) when this method was applicable. Edwards' test is inappropriate if the sample size is less than 50 (Hewitt et al, 1971) or if the seasonal data do not fit a harmonic curve. The fit of the data to a simple harmonic curve was therefore tested (Walter & Elwood, 1975), and where the data were significantly non-harmonic or the sample size was less than 50 the non-parametric method of Hewitt et al (1971) Tables III & IV show the results of testing the seasonality of presentation of the skin cancers, by histology and site, for females and for males.…”

Section: Methodsmentioning

confidence: 99%

“…Edwards' test is inappropriate if the sample size is less than 50 (Hewitt et al, 1971) or if the seasonal data do not fit a harmonic curve. The fit of the data to a simple harmonic curve was therefore tested (Walter & Elwood, 1975), and where the data were significantly non-harmonic or the sample size was less than 50 the non-parametric method of Hewitt et al (1971) Tables III & IV show the results of testing the seasonality of presentation of the skin cancers, by histology and site, for females and for males. Where a cell in these tables is based on less than 50 cases or the data did not fit a harmonic curve, the results of non-parametric testing of the significance of seasonality and the season of peak incidence (taken as the six-month period giving the highest ranking score) are shown in brackets, and the amplitude of seasonality, for which no non-parametric test was available, is not given.…”

Section: Methodsmentioning

confidence: 99%

See 2 more Smart Citations

Seasonality of presentation of cutaneous melanoma, squamous cell cancer and basal cell cancer in the Oxford Region

Swerdlow¹

1985

Br J Cancer

View full text Add to dashboard Cite

Summary The seasonality of presentation of 1019 skin melanomas in Oxford Region 1952Region -1975 Several recent studies (Fears et al., 1977;Houghton et al., 1978;Wigle, 1978;Swerdlow, 1979;Houghton et al., 1980;Houghton & Viola, 1981;MacKie & Aitchison, 1982) have suggested that cutaneous malignant melanoma incidence may have a short induction period after exposure to sun radiation. One finding which could be interpreted as supporting this hypothesis is the seasonal pattern of first diagnosis of melanoma found in one or both sexes in Sweden (Malec & Eklund, 1978), the Third National Cancer Survey (TNCS) areas of the United States (Scotto & Nam, 1980), Hawaii (Hinds et al., 1981) and Western Australia (Holman & Armstrong, 1981), and perhaps in Lane County, Oregon (Morton & Starr, 1979) -such seasonality of presentation could reflect real seasonality of incidence, which could most readily be explained by a short induction period from a seasonal aetiological insult such as ultraviolet radiation exposure. However, several alternative explanations of the seasonal presentation are possible and have not been fully explored in previous work: the disease might be noticed more rapidly in summer and hence present more often then; or the way in which medical care is organised or used might

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Seasonality of presentation of cutaneous melanoma, squamous cell cancer and basal cell cancer in the Oxford Region

Swerdlow¹

1985

Br J Cancer

View full text Add to dashboard Cite

show abstract

“…One interesting aspect of this approach is the ability to incorporate prior knowledge (for example, window lengths based on the climatological length of rainy season at the location) in analyses. Hewitt et al [1971] developed seasonality test in the monthly data based on maximum rank-sum observed for any six-consecutive period. Later, Rogerson [1996] generalized this test, where the peak period can be 3, 4, or 5 months.…”

Section: Seasonality Of Synthetic Datamentioning

confidence: 99%

“…We also used a nonparametric test based on work of Hewitt et al [1971] and Rogerson [1996] to test the null hypothesis of no seasonality against the alternative hypothesis of seasonality in the data. One interesting aspect of this approach is the ability to incorporate prior knowledge (for example, window lengths based on the climatological length of rainy season at the location) in analyses.…”

Section: Seasonality Of Synthetic Datamentioning

confidence: 99%

Nonstationarity in seasonality of extreme precipitation: A nonparametric circular statistical approach and its application

Dhakal

Jain

Gray

et al. 2015

Water Resources Research

View full text Add to dashboard Cite

Changes in seasonality of extreme storms have important implications for public safety, storm water infrastructure, and, in general, adaptation strategies in a changing climate. While past research on this topic offers some approaches to characterize seasonality, the methods are somewhat limited in their ability to discern the diversity of distributional types for extreme precipitation dates. Herein, we present a comprehensive approach for assessment of temporal changes in the calendar dates for extreme precipitation within a circular statistics framework which entails: (a) three measures to summarize circular random variables (traditional approach), (b) four nonparametric statistical tests, and (c) a new nonparametric circular density method to provide a robust assessment of the nature of probability distribution and changes. Two 30 year blocks (1951-1980 and 1981-2010) of annual maximum daily precipitation from 10 stations across the state of Maine were used for our analysis. Assessment of seasonality based on nonparametric approach indicated nonstationarity; some stations exhibited shifts in significant mode toward Spring season for the recent time period while some other stations exhibited multimodal seasonal pattern for both the time periods. Nonparametric circular density method, used in this study, allows for an adaptive estimation of seasonal density. Despite the limitation of being sensitive to the smoothing parameter, this method can accurately characterize one or more modes of seasonal peaks, as well as pave the way toward assessment of changes in seasonality over time.

show abstract

VATERL: An epidemiologic analysis of risk factors

Rittler¹,

Paz

Castilla

1997

Am. J. Med. Genet.

View full text Add to dashboard Cite

This work analyzed the incidence of risk factors in 138 cases presenting two or more of five components defining VATERL, with no other recognized unrelated anomalies: vertebral anomalies, anal atresia, esophageal atresia with or without tracheoesophageal fistula, renal anomalies, and preaxial defects of the upper limbs, including polydactyly of the thumb. The 138 infants were ascertained among 1,811,461 births examined in the 1967-1994 period by the Latin-American Collaborative Study of Congenital Malformations: ECLAMC. One healthy and one malformed control newborn infant were matched to each VATERL case. The birth prevalence rates (per 100,000 births) for VATERL were significantly lower in Venezuela (3.1) than in the other eight countries (8.8) (P < 0.001). Venezuela also had lower rates for all five VATERL defects, even after excluding the 138 VATERL cases. VATERL cases were preferentially males (male proportion 0.6261) (P < 0.02), and, when compared with healthy controls, they had a higher perinatal mortality rate (63.7%) (P < 0.005), a higher frequency of fetal losses in previous pregnancies (12.6%) (P < 0.05), and lower mean birthweights (2,361.79 +/- 809.63 g) (P < 0.005). VATERL cases showed a higher rate than matched malformed controls for prenatal exposures to drugs and physical agents (P < 0.02 and P < 0.05, respectively), although no specific pharmacological or physical group was involved. The lower birth prevalence rates found in Venezuela, for VATERL as well as for each of the five congenital anomalies involved in this association, seem to be biologically meaningful. Since we could not identify a potential risk factor, nor a common cause of underascertainment unique to the Venezuelan subsample and common to all six hospitals, no hypothesis can be advanced here for this phenomenon. Nevertheless, this unequal geographic distribution strongly suggests a common etiopathogenicity for the five congenital anomalies involved in the VATERL association.

show abstract

On Edwards' criterion of seasonality and a non-parametric alternative.

Cited by 69 publications

References 2 publications

Seasonality of presentation of cutaneous melanoma, squamous cell cancer and basal cell cancer in the Oxford Region

Seasonality of presentation of cutaneous melanoma, squamous cell cancer and basal cell cancer in the Oxford Region

Nonstationarity in seasonality of extreme precipitation: A nonparametric circular statistical approach and its application

VATERL: An epidemiologic analysis of risk factors

Contact Info

Product

Resources

About