Time Trend Effects Research Articles

The Effect of Linear Time Trends on the KPSS Test for Cointegration

The so‐called KPSS test for the null hypothesis of cointegration builds on residuals from single equation regressions. Critical values have been provided for regressions with and without detrending. Here it is shown that the latter are not appropriate if the series display linear trends, although this does not mean that detrending is required. In this paper adequate percentiles are suggested for series that follow linear time trends, and tests are based on regressions without detrending. These percentiles are readily available from the literature.

The Search for Causes in an Era of Crime Declines: Some Lessons from the Study of New York City Homicide

This article addresses the problem of testing the effects of particular policies on crime rates in an era of general down trends. One illustration of that problem is our recent finding that rates of non-gun homicide had been declining substantially in New York City for 8 years prior to any significant change in policing and could not plausibly be caused by these later events. The article contrasts three different “controls” for time trend effects, naive cross-sectional controls, detailed models of crime causation, and qualitative checks that examine whether the details of crime patterns are changing in ways consistent with theories of policy events as change agents. The qualitative approach is embraced as a necessity. A final section questions whether criminal justice policies should be assumed to affect general crime rates in broad and undifferentiated ways.

Respiratory diseases in children and outdoor air pollution in São Paulo, Brazil: a time series analysis

OBJECTIVESTo investigate the short term effects of air pollution on the respiratory morbidity of children living in São Paulo, Brazil, one of the largest cities in the developing world.METHODSDaily counts...

Air pollution and hospital admissions for respiratory and cardiovascular diseases in Hong Kong.

OBJECTIVE: To investigate short term effects of concentrations of pollutants in ambient air on hospital admissions for cardiovascular and respiratory diseases in Hong Kong. METHODS: Retrospective ecological study. A Poisson...

Increasing age is a risk factor for psychosis in the elderly.

We examined the association between ageing and the administrative incidence rate of late onset (after age 59) non-organic, non-affective psychosis in two samples of patients aged 60 years or older who were first admitted to hospital in (1) The Netherlands between 1978 and 1992 (n = 8010) and (2) nine regional health authorities in England and Wales (n = 1777) between 1976 and 1978. There was a linear trend in the association between increasing age and first admission rates for non-organic, non-affective psychosis in the elderly, after adjustment for the possible confounding effects of time trend and gender, corresponding to an 11% increase in the incidence with each 5-year increase in age. These observations support a connection between degenerative brain processes and onset of non-affective psychosis in the elderly.

The use of CPI for tourism prices in demand modelling

The non-fare components or tourism prices are often proxied by consumer price indices (CPI) in tourism demand models. The argument for this is that tourism prices move in close concert with general consumer prices. It is not immediately obvious that this is true, as tourist spending differs from general consumer spending. This paper investigates the evidence for the use of CPI for tourism prices, employing a variety of methods and data. For 10 important tourist destinations price series for major tourist expenditure items were estimated. These were found, with a few exceptions, to correlate very highly with the destination's CPI. The high correlations persisted even after linear time-trend effects were removed from the series. In the case of Australia detailed data on tourist origins, numbers, spending breakdowns and on price changes were used to derive sound tourism price indices for major origins of tourists to Australia. The tourism price series were found to correlate very highly with the Australian CPI. The results confirm that it is reasonable to use CPI as a proxy for tourism prices in demand models. However, interpretation of the estimated coefficient is problematical.

Data-dependent allocation rules a comparative study of some for bernoulli data

A clinical trial setting is considered in which two treatments are available for a particular ailment. The response to treatment is Bernoulli with “success” or “failure” as the outcome. Several allocation rules which are designed to reduce the number of patients who receive the inferior treatment are compared with random allocation using simulation. Particular attention is paid to the bias and variance for estimation of the true treatment difference. The effect of time trends in the data is examined.

An analysis of the Japanese expenditure pattern: An application of a dynamic linear expenditure system to the 1961–1982 data

We estimated a dynamic linear expenditure system for 22 subgroups of commodities using a time series of the Japanese family budget data for the last 22 years, and analyzed the factors affecting the changing demands. The proxy for changing tastes was specified as the square of time, taking into account accelerated effects of time trend on expenditures. It would reflect an acceleration of modernization in the mode of living. A demand matrix was established and the effect of changing tastes on each subgroup was measured in an annual rate of shift in demand. Policy implications of the results are discussed in some detail to obtain the long-term prospects for the supply of food commodities in particular.

Time-trend analysis of plasma 11-desoxycorticosterone, corticosterone, cortisol, and aldosterone in fetal and maternal sheep during the last 18 days of gestation

We examined the time trend effects of approaching parturition on fetal and maternal concentrations of cortisol, corticosterone, 11-desoxycorticosterone and aldosterone in eight chronically catheterized pregnant sheep preparations. Fetal plasma cortisol describes a continuously incrementing, straight log-linear function from at least 18 days prior to delivery. Fetal plasma corticosterone describes a time trend effect similar to that of fetal plasma cortisol. Fetal plasma 11-desoxycorticosterone shows no overall time trend effect except for a terminal rise beginning at approximately delivery — 2 days. Maternal plasma cortisol, corticosterone, and 11-desoxycorticosterone concentrations show no overall time trend effects except for similar terminal rises beginning at approximately delivery — 3 days. Both fetal and maternal plasma aldosterone concentrations show wide random fluctuations with no discernible time trend effects. Concentrations of fetal and maternal 11-desoxycorticosterone and aldosterone respectively do not correlate with one another ( P > 0.05) or with cortisol and corticosterone ( P > 0.10) and must therefore be regulated through different control mechanisms. Maternal cortisol, corticosterone, 11-desoxycorticosterone and aldosterone concentrations bear no apparent relationship over time to their fetal counterparts with the approach of parturition.

Interrelationships of Circulating Maternal Steroid Concentrations in Third Trimester Pregnancies. II. C18and C19Steroids: Estradiol, Estriol, Dehydroepiandrosterone, Dehydroepiandrosterone Sulfate, Δ5-Androstenediol, Δ4-Androstenedione, Testosterone, and Dihydrotestosterone

This report describes aggregate time trend effects of advancing gestational age on circulating maternal concentrations of 17beta-estradiol (E2), estriol (E3), dehydroepiandrosterone (D), dehydroepiandrosterone sulfate (D-S), delta 5-androstenediol (delta 5 diol), delta 4-androstenedione (delta 4 A), testosterone (T), and dihydrotestosterone (DHT) in a sequential series of 155 blood samples obtained from 19 normal pregnant women ranging from 26-40 weeks gestational age. Only E2, E3, and D-S show aggregate time trend effects. Log (E2) plots as a linear positive sloping curve from 26-40 weeks. Log (E3) plots as a positive sloping curve that is significantly steeper than log (E2) (P less than 0.05). Log (D-S) plots into a negative sloping curve which mirrors the pattern for log (E2) but cannot be statistically associated with log (E2) except for the opposite sign of their slopes, which are both significantly different from a zero slope (P less than 0.05). delta 4 A, T, DHT, delta 5 diol, and D show no aggregate time trends; however wide, comoving undulations for delta 4 A, T, DHT, and delta 5 diol between 26-28 and 38-40 weeks are confirmed in time by comparison of log mean plots and in magnitude by regressing the C19 steroids on one another. D shows virtually no association with the other C19 steroids. All C19 steroids, except for T, circulate at nonpregnant concentrations, implying that there is little placental secretion of these steroids into the maternal circulation.

Interrelationships of Circulating Maternal Steroid Concentrations in Third Trimester Pregnancies. I. C21Steroids: Progesterone, 16α-Hydroxyprogesterone, 17α-Hydroxyprogesterone, 20α-Dihydroprogesterone, Δ5-Pregnenolone, Δ5-Pregnenolone Sulfate, and 17-HydroxyΔ5-Pregnenolone

This report describes aggregate time trend effects of advancing gestational age on circulating maternal concentrations of progesterone (P), 16α-hydroxyprogesterone, 17α-hydroxyprogesterone (17P), 20α-dihydroprogesterone (20P), Δ5-pregnenolone sulfate, Δ5-pregnenolone (Δ5P), and 17-hydroxyΔ5-pregnenolone (17Δ5P) in a sequential series of 153 blood samples obtained from 19 normal pregnant women ranging from 26-40 weeks gestational age. After logarithmic transformations and curve-fitting procedures, significant differences in slopes were sought and tested by multivariate least squares methods. All steroids, except for 17Δ5P, were markedly elevated above previously reported nonpregnancy concentrations and plotted as positive sloping curves. Log (P) and log (20P) concentration-time curves describe single segment linear curves of similar slope (P > 0.05) and are consistent with the hypothesis that 20P is a metabolite of P, sharing in a relatively stable interconversion with P as the major placental secretory pr...

AGE–MONTH ADJUSTMENT OF CANADIAN DAIRY PRODUCTION RECORDS

Records of Performance (ROP) in dairy production for Ayrshire, Guernsey, Holstein and Jersey breeds were analyzed to estimate constants (age–month subclass averages adjusted for herd differences, time trend and cow culling effect) which can be used for a new Breed Class Average system for simultaneous adjustment of age and month of calving differences. Age, month, and age by month interaction effects were highly significant and thus we conclude that a joint adjustment for age and seasonal variation is necessary. Seasonal variation was sufficiently continuous to warrant age adjustment within individual month. Within breed adjustment for age–month of calving is necessary, but within geographical region and/or lactation number adjustment is not recommended.

Factor Changes and Linear Trends in Eight-Run Two-Level Factorial Designs

Eight-run two level factorial and fractional factorial designs are examined from two points of view: (1) the number of levels which must be changed in performing the design, (2) the effect of a first order time trend on the main effects. It is found that only a few run orders are desirable from these viewpoints and thus randomization of the runs is likely, in general, to lead to an unfavorable sequence.

Effect of Age on Fertility of Bulls in Artificial Insemination

Summary The effect of age on changes in fertility of Holstein and Guernsey bulls was examined in the 60–90 day nonreturn records of an artificial breeding stud. Bulls used for 1,000 or more services per year in the period from 1947 through 1959 were included. Data from the terminal year of service were excluded for each bull culled either because of low fertility or poor semen quality. This exclusion eliminated records on bulls whose fertility dropped markedly in a short period of time. Ninety-one Holstein and 66 Guernsey bulls had one or more records which met the criteria established for inclusion in the analysis. The two groups contributed 322 and 218 bull-years, respectively. In each year, the deviation of the nonreturn rate of each bull from the stud average in percentage points for his breed was calculated. These deviations were used in the analysis of age effects with the intention of reducing the effect of time trends in fertility. Analysis of variance indicated a highly significant difference among age groups (P