Sample Research Paper on SQL
Research Paper Sample on SQL
Although many varieties of sophisticated software are available for database construction, they impose very little restriction on their implemented form. Indeed, it would fly in the face of database philosophy to impose the form of implementation, for this would not encourage features like responsiveness to new requests (Bailey, Creel, Grossman, Gutti, and Sivakumar, 55-57). Thus each operational database is unique, even if the software used in constructing it is an industry standard. A piece of advice one gets from software manuals is to develop two mental pictures of one’s unique database. The first picture is of the contents of the database and the way it is organized; and the second picture is of what can be done with the contents.
Briefly, in the case of the small firms database, the contents were the qualitative and quantitative data obtained with the three instruments AQ 1985, SSI 1985, and RIQ 1988; and the way in which they were handled was governed by a language made available with the software used, known as SQL (Structural Query Language) (Grossman, Hornick, and Meyer, 59-61). This paper starts by analyzing SQL versions as to how it was designed and structured and continues by explaining how SQL may be applied to solve concrete problems and tasks.
The database was initially created in 1985 and then extended in 1988. In 1985 the numerical and categorical data collected for each SBE using the AQ 1985 were stored in one record type (type 1 record) and corresponded to the replies recorded in the questionnaire completed with each owner-manager. Every SBE had one of these records, although information on all items for each firm was not always available. Such data were assigned a missing value. For some of the SBEs there were in addition a set of agenda records which contained the textual information of the summary notes constructed after the SSI 1985 (Grossman, Creel, Mazzucco, and Williams, 115-120). Each firm was uniquely identified as a ‘case’ in the database. An additional dummy firm, number 99, was used as a technical device to link the agenda records containing the probes of the SSI 1985 (Bailey, Creel, Grossman, Gutti, and Sivakumar, 55-57).
In 1988 two further levels (i.e. records types) were added to the database, one containing numerical and categorical data from the RIQ 1988 (type 3 record) and the other containing textual data from the RIQ 1988 (type 4 records). Thus the final form of the database involved seventy-three firms treated as cases, with four record types attached to each firm (or case), two record types containing numerical and categorical data and two record types containing textual data. For convenience, this paper refers to the database as having four levels, each level referring to a type of record (Grossman, Hornick, and Meyer, 59-61).
At the time of updating and extending the database in 1988, the opportunity was taken to undertake a full audit of the data mounted in 1985 by reference to the original physical interview schedules. In the autumn of 1989 a further full audit was undertaken of the data added in 1988, again by reference to the original schedules. Some design modifications were also implemented in autumn 1989 to enhance the compatibility of records. As a consequence one hopes that the data, whilst not perfect, are of about as high a quality as the usual human and computing resources will permit.
Once fully installed, the database contained 29,200 data items in the AQ 1985; 2,635 data items in the SSI 1985; and 3,519 data items in the RIQ 1988 making 35,354 data items in all. Allowing a mean value of 4 characters per data item the total disk space used was 141.416 k bytes.
The data are simply a set of tables. These tables may be manipulated by an algebra or a logical calculus. However, on the grounds of simplicity, a language which is as similar to English as possible is to be preferred. The best known of such languages is SQL(Structured Query Language) first developed by IBM for internal purposes, but now used much more widely. The full ‘vocabulary’ is rich and extensive, but here it is necessary to do no more than illustrate the use of a very limited part of it.
The basic building block of the SQL language is the ‘select’ command. For example, if numerical and categorical data on firms in 1985 are gathered in the type 1 record of the database relating to the AQ 1985, and denoted as level 1, a select statement might take the form: select firm sic employ trainno from 1 where employ le 2 (Grossman, Creel, Mazzucco, and Williams, 115-120)
One is interested here in firms at the very lower end of the size distribution, for which employment is less than or equal to (i.e. ‘le’) 2 persons. For such firms, one wants to know whether they take on trainees, and if so, how many. Further, one wants to know which industrial classification (SIC code) each of these SBEs lies within. This sort of information might be relevant to an analysis of training functions in an enterprise culture. It is possible to report, having implemented the above instruction, that twenty-six firms (35 per cent) employed two or less full-time workers.
The database also provides all the requested SICs and full information on trainees, but they are too lengthy to report upon in detail in this illustration. Briefly, just four of these firms with less than three employees also employ trainees. That is just 5 per cent of the total and only 15 per cent of this sub-sample of very small firms. Two of these SBEs were in manufactures and construction, and two were in transport and services. By contrast, 47 per cent of firms employing more than two persons did also employ trainees, the great bulk (82 per cent) of these being in manufactures and construction. Here, the average number of trainees was two. Broadly, the training function in this sample tends to be most fostered in the larger SBEs in manufactures. To pick out those specific firms which employed two or less full-timers and had trainees (irrespective of numbers), one would write the instruction, select firm from 1 where employ le 2 and train eq 1 (Bailey, Creel, Grossman, Gutti, and Sivakumar, 60-64)
Here, train=1 is the value of the binary variable when the firm does employ trainees. Such binary variables are a common feature of the database.
Other refinements are possible. As suggested above, one could look at the distribution of the SBEs across industrial classifications. SICs of 60 or over relate to transport and services. Thus the instruction:
select firm sic from 1 where employ le 2 and sic ge 60 (Bailey, Creel, Grossman, Gutti, and Sivakumar, 60-64) gives us the number of SBEs in transport and services that employ less than three people. There were twelve of these (46 per cent of the sub-sample for which employment ≼2). Further, one can focus on more qualitative issues. Referring now to the RIQ 1988, one might ask whether the owner-managers of such firms sensed the emergence of an enterprise culture in Scotland since 1985. To be even more specific, what was the reaction of an SBE in, say, wholesale distribution (SIC=61) and an SBE in, say, recreational services (SIC=97)? Level 1 of the database informs us that two firms were in the relevant size range and operating in wholesale distribution in 1985. Level 4 of the database informs us that one of these owner-managers thought that the government had ‘led the way’ and that more people would ‘have a go on their own’. This owner-manager observed that older SBEs had better subsequent survival rates than younger SBEs.
Method of Experimentation
According to level 3 of the database the other owner-manager in wholesale distribution (whose SBE distributed cane- and basket-ware) had ceased trading by 1988 and could not be traced. There were also two firms in the sub-sample operating in recreational services in 1985 (Grossman, Creel, Mazzucco, and Williams, 120-123). One owner-manager, a tour operator, thought the successful setting up of many small firms selling tours in recent years was indicative of an emerging enterprise culture. The other SBE in recreational services had ceased trading by 1988, though a different owner-manager was running a different business from the same premises in 1988. All the above qualitative information was extracted from level 4 (the type 4 records) in the database which held text from the RIQ 1988 (Grossman, Creel, Mazzucco, and Williams, 120-123).
Pursuing qualitative evidence further, one could extract text from the type 2 records (i.e. level 2 of the database) which related to the probes of the SSI 1985 (Chapman, Clinton, Kerber, Khabaza, Reinhertz, Shearer, and Wirth, 23-27). For example, for the firm in wholesale distribution that survived, a detailed account was given of the nature of rivalry. The SBE was a wholesale wine business. The owner-manager felt the increased pressure of competition mounting ‘day by day’. A particular pressure was felt from brewers who had diversified into wine sales because wine had become fashionable and beer sales were on the wane. It was felt that profit margins of SBEs in wine distribution were being squeezed by these new big players in the market. Balanced against this, in 1985 wine sales were growing at 24 per cent per annum in the UK and were expected to continue at this rate for two further years, with market saturation being reached after five years. It was felt that cheap continental holidays had given customers favorable consumption experience of wine, which had subsequently encouraged the boom in the UK market. Brewers who had diversified into wine had high strategic stakes and were here to stay.
Results of the experimentation
For the firm in recreational services that survived, the tour operator, a similarly full account of rivalry was given. The fact that many SBEs operated in the business was explained by low fixed costs and opportunities for specialization. The number of firms was continuing to rise because of perceived growth potential in Scotland. The location was felt to be well suited to touring both because of Scotland’s natural beauty and because an extensive infrastructure, which tour operators could readily utilize, was already in place. It was observed that there was as much specialization amongst tour operators as there were kinds of holidays (Bailey, Creel, Grossman, Gutti, and Sivakumar, 60-64).
Seasonality meant that there was intermittent overcapacity, but at the time of interview in 1985 the owner-manager was experiencing excess demand for his firm’s services and was sub-contracting. He felt that the industry was robust, and markets were so highly fragmented that SBEs could always find niches that did not overlap with other firms’ markets. This clustering of products in tours can be explained by the information externality analysed by Pepall (1990). Firms can observe information about demand by looking at the market for similar products produced by rivals. At low cost they can market new goods in those parts of characteristics space which are ‘spanned’ by the qualities of existing products (over characteristics like transport, meals, accommodation, and entertainment).
Both the examples on rivalry are of firms that survived to 1988 and which operated in buoyant markets. One could go on to enquire why these two firms survived but their partner firms in the same SIC classifications did not. The failed firm in wholesale distribution did not complete the SSI 1985, but the failed firm in recreational services did. The owner-manager in the latter case ran an SBE which provided ‘kit’ airplanes for aerobatics. In 1985, looking at the agenda item on rivalry, one notes the owner-manager expressing doubts about his product and the market. It depended on the existence of a special type of customer who had ‘flying in his blood’ and was ‘very self-indulgent’.
Investment was over a long period of time and during that time there was no return; only four kits could be constructed a year; and government regulations on components and construction were very strict. Clearly, survival for this firm was not going to be easy. One can probe further, using two good predictors of survival: the gearing ratio and cash flow. Higher gearing lowers the probability of survival, as does the existence of cash flow problems (Bailey, Creel, Grossman, Gutti, and Sivakumar, 55-57).
Accessing level 1 of the database, using select commands similar to those above, one discovers that of the two firms in recreational services both had experienced cash flow difficulties, but the one which failed (the kit aircraft firm) was much higher geared (250 per cent compared to 75 per cent) and arguably, therefore, more exposed to risk. Looking at the two firms in wholesale distribution, the one that survived (the wine distributor) had not experienced cash flow problems, whereas the one that failed (the cane- and basket-ware distributor) had. Further, the one that survived had a zero gearing (i.e. no debt at all), whereas the one that failed had a 400 per cent gearing ratio. One could go on and make the investigations increasingly complex, looking at advertising form, advertising strategy, assets, and so on, but perhaps enough has been done to demonstrate the methodology.
The sorts of data illustrated above are very diverse, and have been presented in terms of tightly defined variables (for example, employment) or well-defined agenda items (for example, rivalry). The significant point to realize is that all the above data, numerical, categorical, and textual, were obtained without departing from a computer terminal. All retrievals were obtained using the SIR software. In a more complicated context, the characterization of the ‘representative SBE’ may be achieved using similar, but obviously much more extensive, database manipulations (Chapman, Clinton, Kerber, Khabaza, Reinhertz, Shearer, and Wirth, 23-27).
Although the small firms research area is very empirically oriented, and a welcome diversity of data types are currently in use, there does not yet seem to have been significant innovation in the way in which data are handled. An enormously impressive aspect of Birch’s influential early work was its handling of large data sets. Yet it impressed by its scale, rather than by its method. A significant strand of argument throughout has been the advocacy of more diverse methods of data collection. Unless these new sorts of data are handled in innovative ways, using the benefits of recent advances in database design, research may not achieve the conceptual breakthrough that those active in the small firms area hope for. Instead, a greater volume of data, in increasingly diverse forms, will simply overwhelm the investigator. Currently quite complex methods are used on very large numerical data sets which are organized in quite strict ways (e.g. in a hierarchical fashion). Data processing methods used might involve the concatenation of files, the selection of sub-sets of data, the definition of new variables from old, and the editing of data into a form which makes them suitable for further analysis using statistical and econometric software.
Remember, free research papers, sample research papers and research paper examples on SQL topics are traced by plagiarism detection systems. All samples online are plagiarized. Don’t download them and submit them as your own research project for high school, college or university. Why not to get a 100% original research paper at PapersMart.net? Need a free quote?
If you need a custom research paper on SQL feel free to contact our online research paper writing company. Our professional academic writers who hold PhD and Master’s degree will write a 100% non-plagiarized research paper, term paper, essay, research proposal or dissertation for you. Our custom research paper service produces high-quality custom papers on any topics and disciplines. On-time delivery and confidentiality guarantee!