Enhancing estimation methods for integrating probability and nonprobability survey samples with machine-learning techniques. An application to a Survey on the impact of the COVID-19 pandemic in Spain

Web surveys have replaced Face-to-Face and computer assisted telephone interviewing (CATI) as the main mode of data collection in most countries. This trend was reinforced as a consequence of COVID-19 pandemic-related restrictions. However, this mode still faces significant limitations in obtaining probability-based samples of the general population. For this reason, most web surveys rely on nonprobability survey designs. Whereas probability-based designs continue to be the gold standard in survey sampling, nonprobability web surveys may still prove useful in some situations. For instance, when small subpopulations are the group under study and probability sampling is unlikely to meet sample size requirements, complementing a small probability sample with a larger nonprobability one may improve the efficiency of the estimates. Nonprobability samples may also be designed as a mean for compensating for known biases in probability-based web survey samples by purposely targeting respondent profiles that tend to be underrepresented in these surveys. This is the case in the Survey on the impact of the COVID-19 pandemic in Spain (ESPACOV) that motivates this paper. In this paper, we propose a methodology for combining probability and nonprobability web-based survey samples with the help of machine-learning techniques. We then assess the efficiency of the resulting estimates by comparing them with other strategies that have been used before. Our simulation study and the application of the proposed estimation method to the second wave of the ESPACOV Survey allow us to conclude that this is the best option for reducing the biases observed in our data.     

杂交试验中的最佳样本含量

杂交试验是一项费时费钱的工作,因此在进行试验之前如能进行严密的设计,给出试验所需的样本大小是十分必要的,统计学中常见的估计样本容理的公式不宜应用于杂交试验,本文分两种情况给出了杂交试验中样本容量的估计公式,据此估计出的样本容量安排杂交试验,可在满足试验者要求的条件下,使试验的总成本最低或使试畜的总头数最少。     

ESTIMATING PUP PRODUCTION OF HARP SEALS, PAGOPHILUS GROENLANDICUS, IN THE NORTHWEST ATLANTIC

Photographic and visual aerial surveys to determine current pup production of Northwest Atlantic harp seals were conducted off Newfoundland and in the Gulf of St. Lawrence during March 1999-Photographic surveys were conducted on all whelping concentrations between 14 and 24 March, whereas a visual survey was made of the southern Gulf concentrations on 14 March. Pup production was estimated to be 739,100 (SE = 96,300, CV = 13.0%) at the Front, 82,600 (SE = 22,500, CV = 27.2%) in the northern Gulf, and 176,200 (SE = 25,400, CV = 14.4%) in the southern Gulf (Magdalen Island) for a total of 997,900 (SE = 102,100, 10.2%). Changes in aerial survey estimates indicate that pup production has increased since 1994. A new method to correct for the temporal change in the proportion of pups present on the ice was examined by fitting the percentage of pups observed in three age-dependent stages to a Normal distribution. The results were compared to those obtained from a more complex model used previously. The Simple model produced slightly higher, and hence more conservative, estimates of the proportion of births that had occurred before the time of the survey than the Complex model. When using the Simple model fewer assumptions regarding the start date of pupping and the proportion of older pups remaining on the ice were required, the herd had to be followed for a shorter period, and a more convenient means of calculating confidence limits was available.     

DIET OF HARBOR PORPOISES IN THE KATTEGAT AND SKAGERRAK SEAS: ACCOUNTING FOR INDIVIDUAL VARIATION AND SAMPLE SIZE

Stomach contents of 112 bycaught harbor porpoises ( Phocoena phocoena ) collected between 1989 and 1996 in the Kattegat and Skagerrak seas were analyzed to describe diet composition and estimate prey size, to examine sample size requirements, and to compare juvenile and adult diets. Although porpoises preyed on a variety of species, only a few contributed substantially to the diet. Atlantic herring ( Clupea harengus ) was the dominating prey species for both juveniles and adults. Our results, in combination with those from previous studies, suggest that where herring is a dominant food source, porpoises prey primarily on size classes containing mature or maturing individuals. Further, we also show that Atlantic hagfish ( Myxine glutinosa ) may be an important food resource, at least for adult porpoises. Examination of sample size requirement showed that, depending on the taxonomic level used to describe the diet, a minimum of 35–71 stomachs are needed to be confident that all common prey species will be found.     

STABILITY IN THE PROPORTION OF HARBOR SEALS HAULED OUT UNDER LOCALLY IDEAL CONDITIONS

We monitored the haul-out behavior of 68 radio-tagged harbor seals ( Phoca vitulina ) during the molt season at two Alaskan haul-out sites (Grand Island, August-September 1994; Nanvak Bay, August-September 2000). For each site, we created a statistical model of the proportion of seals hauled out as a function of date, time of day, tide, and weather covariates. Using these models, we identified the conditions that would result in the greatest proportion of seals hauled out. Although those "ideal conditions" differed between sites, the proportion of seals predicted to be hauled out under those conditions was very similar (81.3% for Grand Island and 85.7% for Nanvak Bay). The similar estimates for both sites suggest that haul-out proportions under locally ideal conditions may be constant between years and geographic regions, at least during the molt season.     

Strategies for microarray analysis of limiting amounts of RNA.

One of the critical limitations of current microarray technologies for use in expression analyses is the relatively large amount of input RNA required to generate labelled cDNA populations for array analysis. In situations where RNA is limiting, the options for expression profiling are to increase cDNA labelling and hybridisation efficiency, or to use an amplification strategy to generate enough RNA/cDNA for use with a standard labelling method. Sample amplification approaches must preserve the representation of the relative abundances of the different RNAs within the starting population and must also be highly reproducible. This review evaluates current signal and sample amplification technologies, including those that can be used to generate labelled cDNA populations for array analysis from as little as a single cell.     

Natural Occurrence ofChondrostereum purpureumin Relation to Its Use as a Biological Control Agent in Canadian Forests

The incidence of natural fructification ofChondrostereum purpureumwas estimated quantitatively on southern Vancouver Island during two winter seasons in randomly located 1000-m²plots and compared with potential added fructification that might occur as a result of using the fungus to control stump-sprouting of hardwood weeds in young forest stands. Fructification was surveyed in forests as well as in urban or agricultural areas by estimating the surface areas of woody substrates covered with basidiocarps. In addition to random plots, estimates were made also in locations where the fungus would be expected to occur (woodpiles, silvicultural thinnings, and killed trees). Basidiocarps were found throughout the area in various types of forest cover as well as in urban or agricultural situations. The amount of added fructification through the use of the fungus as a biological control agent was determined from inoculated plots as well as from calculated stump-surface areas developed from published stand-density data. Added fructification was multiplied by a factor representing the maximum biological control frequency in order to compare added fructification with natural fructification values. From the various calculations, it was determined that the added fructification ofC. purpureumis of the same order of magnitude as naturally occurring levels or even lower. In addition, there is a distinct geographical separation between predominantly forestry and predominantly settled areas where fruit and ornamental trees are cultivated. Accordingly, it was concluded that using the fungus as a biological control agent in forestry is not likely to pose a significant threat to fruit growing and commercial forests.     

On-line immunoanalysis of monoclonal antibodies during a continuous culture of hybridoma cells

The monoclonal-antibody production of an immobilized hybridoma cell line cultivated in a fluidized-bed reactor was monitored on-line for nearly 900 h. The monoclonal antibody concentration was determined by an immuno affinity-chromatography method (ABICAP). Antibodies directed against the product, e.g. IgG, were immobilized on a micro-porous gel and packed in small columns. After all IgG present in the sample was bound to the immobilized antibodies, unbound proteins were removed by rinsing the column. Elution of the bound antibodies followed and the antibodies were determined by fluorescence. The analytical procedure was automated with a robotic device to enable on-line measurements. The correlation between the on-line determined data and antibody concentrations measured by HPLC was linear. A sampling system was constructed, which was based on a pneumatically actuated in-line membrane valve integrated into the circulation loop of the reactor. Separation of the cells from the sample stream was achieved by a depth filter made of glass-fibre, situated outside the reactor. Rapid obstruction of the filter by cells or cell debris and contamination of the sample system was avoided by intermittent rinsing of the sample system with a chemical solution. The intermittent rinsing of the filter, which had a surface of 4.8 cm², resulted in an operational capacity of up to 40 samples (1.0 l total sample volume). Both the sampling system and the analytical device functioned without failure during this long-term culture. The culture temperature was varied between 34 and 40 °C. Raising the temperature from 34 up to 37 °C resulted in a simultaneous increase of growth and specific antibody production rate. Specific metabolic rates of glucose, lactate, glutamine and ammonium stayed constant in this temperature range. A further enhancement of temperature up to 40 °C had a negative effect on the growth rate, whereas the specific monoclonal antibody production rate showed a small increase. The other specific metabolic rates also increased in the temperature range between 38 to 40 °C. This revised version was published online in July 2006 with corrections to the Cover Date.     

A survey of black dot and other fungal tuber blemishing diseases in British potato crops at harvest

Black dot and other fungal blemishing diseases were assessed on tubers from crops in eastern England in 1987 and 1988, and in 1989 and 1990 from crops throughout Great Britain. Black dot was found on tubers in crops from all areas of the country, and was most common in those from eastern and southern England. The disease was scarce on most Scottish crops, but occasionally some were severely affected. In 1988 there was no consistent relationship between the amounts of disease on the seed and subsequent ware crop. On average, the disease was more prevalent in irrigated crops and where the interval between successive potato crops was short. In 1990 treating seed tubers with fungicide had no effect on subsequent levels of disease. Few crops were unaffected by silver scurf, and it was most common in crops from southern and eastern England. In the national survey common scab was the most prevalent disease, black scurf was less common and skin spot and powdery scab uncommon except in some Scottish crops.     

Fungal spores are an important component of library air

The airborne fungal spore types were studied in different libraries in Delhi, using an Andersen sampler and a Burkard personal sampler, for culturable and non-culturable fungi respectively. The concentration inside the libraries, before and after the agitation of books, were compared with outside air. The major contributors to the library air areCladosporium, aspergilli/penicillia, smuts andAlternaria, varying from 50 to 14%. Some fungi (Cladosporium, Alternaria, smut,Penicillium chrysogenum andnigricans) showed seasonal occurrence, corresponding to their occurrence in the extramural environment. Aspergilli/penicillia,Drechslera, Curvularia andAspergillus flavus had a significantly higher concentration (P<0.01) inside the library, and recorded a significant increase in concentration after agitation of books. Air-conditioned libraries have low fungal spore concentrations, as compared to naturally ventilated libraries.