Methodological issues. Big Data can be used in several ways when producing official statistics. It can be used as the main source of information, as an additional data source to aid estimates primarily based on survey or administrative data or merely to improve the weights in model-based estimation or to impute missing values. In all cases the methods used may differ. As a consequence, each of the issues described below could be interpreted from these different points of view. It was therefore decided to consider the experiences in WPs 1 to 7 of the ESSnet as the starting point for the focus of each section. Hence, the focus is on methodology when using Big Data in practice. The experience gained here, forms the basis of a -to be developed- general applicable Big Data methodology.
2.1. What should our final product look like?
Methodological issues. Avoiding Pitfalls where Possible So far this introduction discussed the issue of penetrability of perception and specified what aspects of penetrability of auditory perception this thesis will focus on. As mentioned above, Firestone and Xxxxxx (2016) listed six pitfalls that research into top-down effects on perception (or the penetrability of perception) should avoid. The current section will discuss these pitfalls and how we took those into account. The pitfalls will be discussed in the order in which they are addressed in the chapters of this thesis. The discussion of these pitfalls will also provide further insight into the methods we have used in our studies and will lead to the introduction of Chapter 5. In addition, this section will provide some comments to the pitfalls that will not be explicitly addressed in the experiments presented in the current thesis. It should be noted that frequently, albeit often implicitly, a distinction is made between perception, judgment, and response (e.g., in Firestone & Xxxxxx, 2016). Even though the existence of such a clear distinction is debatable and depends on one’s view of how the mind is organized (e.g., Xxxxx, 2013; Xxxxxxxx, Xxxxxxx, & Xxxxxxxxxxx, 2008; Xxxxxxxx & Xxxxxx, 2007), to follow Firestone and Xxxxxx (2016), the discussion of the first two pitfalls will be organized according to this distinction. First, the next section, on demand and response bias effects, will address the possible discrepancy between judgment and response that may confound results (pitfall #3 in Firestone & Xxxxxx, 2016). Second, the subsequent section, on measuring effects on judgment rather than perception, will delve into the distinction between perception and judgment (pitfall #2 in Firestone & Xxxxxx, 2016). According to Xxxxxxxxx and Xxxxxx (2016), many findings that have been interpreted as cognitive or affective penetration of perception may instead be explained as effects of demand characteristics, and particularly of task demands. Other response biases may also lead to a discrepancy between the actual judgment and the response that participants provide. The sections below will first focus on demand characteristics and will then address several other response biases and how these can be avoided or controlled. Demand characteristics are cues, for example from the experimental context, experimenter, or task, that inform the participant about the purpose and hy- pothesis of the experiment and that influence how the parti...
Methodological issues. In terms of methodological issues, the contractor should make recommendations about a number of issues. These include: the optimal frequency of data matching for Next Step customer data, the optimal time period for looking at the impact of interventions, the options for identifying a counterfactual / comparison group in future analysis and any further feasible avenues for analysis.
Methodological issues. This study has sought to identify the additional savings or costs that would accrue to a company if there were certain changes in the accounting disclosure requirements of the Directive. As such, the questions used to gather data have asked for the change in the amount of time spent by a company or its advisers, and changes in direct expenditure. This information has been requested both for the first year – when there might be set up expenditure – and for subsequent years. Company and professional adviser time has been costed using cost rates discussed later in this report The resulting quantified costs or savings represent the marginal cost or saving from the change in the disclosure requirement. In other words, the cost or saving is the change in the cost of accounts preparation from the proposed change in the Directive. What the study has not done is to try to quantify the whole of the cost of the accounting system. The study has assumed that the basic accounting system remains unchanged and that all that changes is the final reporting requirement for statutory accounts. In practice, this may well be a small proportion of the total cost of accounting. In analysing the cost information provided by companies, it became apparent that a small number of companies said that there would be large costs or savings from some of the changes. These outlying values had a significant effect on the arithmetic mean of costs, so for analysis purposes the median of response has been used. Both mean and median values are shown in the detailed analysis in appendix B. Appendix H contains details of how medians were estimated.
Methodological issues. This work employed two methodological approaches. Firstly, we used the conceptual approach to address research questions and issues tackled by this work. Secondly, we used collection of primary data from the Retailers in Multan which is a city of Pakistan. The data collection was done by using a survey to retailers called ”Kiryana Stores” and ”General Stores” in Pakistan, where ”Kiryana Stores” are traditional small shops where goods are mostly stocked in sacks kept on the floor. These shops use over 50 % of the floor space for unbranded grocery items such as spices, sugar, pulses, rice, floor, and kerosene oil. These shops also store some branded and packaged items like dairy products, personal care products, and traditional fast food. They are typically 300 square feet in size and ”General Stores” typically allocate fifty percent shelf space to branded and packaged category grocery items including foods, toiletries and household goods. Products are generally stocked on the floor or wall-mounted shelves (Government of Pakistan, Survey of Wholesale and Retail Trade, 1998). They are mostly high frequency stores where volume of sales per customer is low but the turnover is high because of high population density. Service is typically provided over the counter, rather than through self-service. They are typically 500 square feet in size. Similarly, use of secondary data was also made wherever necessary from different sources which provided basis for this research.
4.3.1 Empirical Setting According to the researchers, a particular industry is chosen to test the domain of a particular theory in particular settings. While testing theory, internal validity is given more importance over external validity is the primary reason for this choice. With selection of a specific industry internal validity will be improved. Similar thinking considers such a choice as a better way to minimize the chances of error and high variance while simultaneously improving statistical and predictive powers of the study. In this regards the retail sector in Pakistan was effectively used to test the current research setting. The retail sector in Multan mainly consists of Kiryana Stores and General Stores or high frequency stores where the retailing has not yet evolved to huge retail chain stores and super markets. Although, as the German retail giant ‘Metro’ is considering now to open up their outlets in various Pakistani cities including Multan, it is expected that the scenario in the r...
Methodological issues. We discuss below two features relevant to the monetary assessment of the mortality impacts associated with the traditional HIA approach.
Methodological issues. As pointed out by Xxxxxxx et al.(2015a), when assessing the net impacts of carsharing, it is necessary to know: how individuals travelled before and what modal behaviours they changed due to carsharing and how individuals would have travelled in the absence of carsharing (e.g., postponed vehicle purchase). Activity data alone cannot answer these questions, and surveys are required. An additional complication is that large (i.e. nationwide) surveys are typically unsuitable for the evaluation of shared mobility because (Xxxxxxx et al.2015a): They do not reflect the dynamics of behavioural change before and after people start using the system; Within the overall sample, the sample of people using shared mobility services is typically small; There are important time lags between subsequent surveys; as a result, they cover different samples of the population. Because of these limitations, Xxxxxxx et al. (2015a) argue in favour of dedicated surveys of members of carsharing schemes. However, as we shall discuss in detail below, such dedicated surveys face problems of their own, such as self-selection of the sample. Xxxxxxxx and Xxxxxxx (2015) discuss in detail two other fundamental methodological problems associated with the empirical assessment of carsharing schemes. The long term impacts of carsharing may be quite different from the immediate impacts, and the transition to a stable situation may take years. In the short term, carsharing may induce new travel as zero-car households start to drive carsharing-cars. In the longer run, households may shed their private cars, which can be expected to lead to a decrease in car usage21. There are no agreed standards for the evaluation of carsharing-systems. As a result, “not a single study-design in the field of empirical carsharing research has ever been replicated”. Moreover, most existing studies are static in nature. There are at least three possible approaches to the evaluation of carsharing impacts: Studies measuring the hypothetical impacts ask carsharing users how they would hypothetically cover their mobility-needs today, if their currently used carsharing- system was not offered. Studies measuring retrospective impacts compare the mobility behaviour before and after people join carsharing-schemes. Studies measuring future impacts ask new carsharing-users for their planned future mobility-changes due to carsharing. Xxxxxxxx and Xxxxxxx (2015) argue that the appropriateness of a method depen...
Methodological issues. This work employs two methodological approaches. The first approach is a conceptual one that aims at addressing research issues tackled by this work. This was done by the use of secondary data. The second approach involved the collection of relevant primary data. This was done through the use of a survey of traders in the value chain of pigeonpea in Tanzania.
Methodological issues. Although the majority of respondents were born and raised in the place where I collected their surveys, the one group for which this was not always the case was the 18- 30 year-olds of Siena. The University of Siena traditionally attracts many students from the southern regions of Italy, and of the respondents who filled out the survey in Tuscany, 42% indicated a region other than Tuscany as their place of birth. Only 25% percent of the surveys collected in Veneto and 13% of the surveys collected in Campania were filled out by respondents born elsewhere. Although a total of 140 out of 530 respondents were not born in the region of data collection, it is important to remember that all of them were residents of that region. The survey was for the most part easily understood and easily completed, but 31+ respondents in particular appeared to find the map-coloring task difficult, either because they felt overwhelmed by the range of dialects to choose from or because they were unsure of their ability to correctly identify all the regions of Italy (or both). Respondents aged 18-30 appeared to find the map-coloring task less difficult—perhaps because they were familiar with fewer dialects and so felt less overwhelmed by choice. For both older and younger respondents, the multiple-choice questions proved unproductive. This was due to poor wording on my part, as well as the limitations inherent in multiple-choice responses. For instance, the question, “Capisci un dialetto anche se non lo parli21” was meant to ask, “What dialect do you understand even if you don’t speak it?” However, many respondents, having already indicated that they did speak a dialect, left this question blank, probably interpreting it as, “If and only if you don’t speak a dialect, which one do you understand?” Additionally, though respondents often gave me detailed verbal answers to the multiple-choice questions, on the survey itself they were forced to choose from written options that were much more limiting. Although a couple of respondents actually wrote free-response-style answers even for the multiple-choice section, in general, much was lost in translation. There was sometimes a tendency of subjects to judge or take issue with my own language, either positively or negatively—comments ranged from “You speak so well for an American!” to “There are mistakes in this survey,” and one middle school teacher corrected my survey even as she completed it. My lack of a readily identifiable regional ...
