Data collection
Research Methodology
Research Methodology is the complete plan of attack on the central research problem. It provides the general structure for the procedures that the researcher follows, the info that the researcher collects, and therefore the data analyses that the researcher conducts, thus involves planning. It’s an idea with the central goal of solving the research problem in mind. Research methodology concerning how the study was held. It includes; research design, study population, sample and sample size, methods of knowledge collection, methods of knowledge analysis, and anticipation of the study. Research methodology concern with a philosophy of research development. It includes the assumptions and values that serve a rationale for research and therefore the standards or criteria the researcher uses for collecting and interpreting data and reaching conclusions (Martin and Amin, 2005:63). In other words research methodology determines the factors like the way to write a hypothesis and what level of evidence is important to form decisions on whether to simply accept or reject the hypothesis.
Data Collection
Data collection is defined because of the procedure of collecting, measuring, and analyzing accurate insights for research using standard validated techniques. A researcher can evaluate their hypothesis on the idea of collected data. In most cases, data collection is the primary and most vital step for research, regardless of the sector of research. The approach of knowledge collection is different for various fields of study, counting on the specified information.
Data collection is the method of gathering and measurement data on variables of interest, in a very old systematic fashion that allows one to answer expressed research questions, check hypotheses and evaluate outcomes. The info collection component of research is common to all or any fields of study including physical and social sciences, humanities, business, etc. While methods vary by discipline, the stress on ensuring accurate and honest collection remains equivalent.
Data collection may be the systematic process of gathering observations or measurements. Whether you are activity research for business, governmental or educational purposes, information collection permits you to comprehend first-hand data and original insights into your research problem. While methods and aims may differ between fields, the general process of knowledge collection remains largely equivalent. Before you start collecting data, you would like to consider:
• The aim of the research
• The sort of data that you simply will collect
• The methods and procedures you’ll use to gather, store, and process the info
The importance of making certain correct and acceptable information collection
Regardless of the arena of study or preference for outlining information (quantitative, qualitative), correct data collection is very important in maintaining the integrity of research. Each the selection of acceptable data collection instruments (existing, modified, or new developed) and clearly represented instructions for his or her correct use reduces the probability of errors occurring.
Consequences from improperly collected information include
• Inability to answer the research questions accurately
• Inability to repeat and validate the study
• Distorted findings resulting in wasted resources
• Deceptive different researchers to pursue unproductive avenues of investigation
• Compromising decisions for public policy
• Inflicting hurt to human participants and animal subjects
While the degree of impact from faulty data collection could vary by discipline and so the character of the investigation, there is the potential to cause disproportionate damage once these research results are wont to support public policy recommendations. Issues related to maintaining the integrity of information collection:
The primary principle for preserving data integrity is to support the detection of errors inside the data collection methodology, whether or not or not or not they are created designedly (deliberate falsifications) or not (systematic or random errors).Most, Craddick, Crawford, Redican, Rhodes, Rukenbrod, and Laws (2003) describe ‘quality assurance’ and ‘quality control’ as two approaches that will preserve data integrity and make sure the scientific validity of study results. Each approach is implemented at different points within the research timeline (Whitney, Lind, Wahl, 1998):
1. Quality assurance - activities that happen before data collection begins
2. internal control - activities that happen during and after data collection
Since quality declaration precedes data collection, its main meeting point is ‘prevention’ (i.e., averting problems with data collection). Prevention is the most cost-effective activity to make sure the integrity of data collection. This practical measure is best verified by the standardization of procedures developed during a broad and detailed procedure manual for data collection. Poorly written manuals increase the danger of failing to spot problems and errors early within the research endeavor. These failures could also be demonstrated in a number of ways:
• Uncertainty about the timing, methods, and identity of the person(s) liable for reviewing data
• Partial listing of things to be collected
• Vague description of knowledge collection instruments to be utilized in lieu of rigorous step-by-step instructions on administering tests
• Failure to spot specific content and methods for training or retraining staff members liable for data collection
• Vague instructions for using, making adjustments to, and calibrating data collection tools (if appropriate)
• No identified mechanism to document changes in procedures that will evolve over the course of the investigation.
An essential component of quality assurance is developing a rigorous and detailed recruitment and training plan. Inherent training is that they got to effectively communicate the worth of accurate data collection to trainees (Knatterud, Rockhold, George, Barton, Davis, Fairweather, Honohan, Mowery, O’Neill, 1998). The training aspect is especially important to deal with the potential problem of staff that may unintentionally deviate from the first protocol. This phenomenon referred to as ‘drift’, should be corrected with additional training, a provision that ought to be laid out in the procedures manual.
Given the variety of qualitative research methods (non-participant/ participant observation, interview, archival, field study, ethnography, content analysis, oral history, biography, retiring research) it’s troublesome to make generalized statements regarding however one should establish a search protocol so as to facilitate quality assurance. Certainly, researchers conducting non-participant/participant observation may have only the broadest research inquiries to guide the initial research efforts. Since the researcher is that the most measurement device throughout a study, repeatedly there are very little or no different information collection instruments. Indeed, instruments may have to be developed on the spot to accommodate unanticipated findings.
While internal control activities (detection/monitoring and action) occur during and after data collection, the small print should be carefully documented within the procedures manual. A clearly outlined communication structure could also be a necessary pre-condition for establishing observation systems. There should not be any uncertainty concerning the flow of data of knowledge of information} between principal investigators and employees members following the detection of errors in data collection. A poorly-developed communication structure encourages lax observation and limits opportunities for detecting errors.
Detection or monitoring can take the shape of direct staff observation during site visits, conference calls or regular and frequent reviews of knowledge reports spotting inconsistencies, extreme values or invalid codes. While site visits may not be acceptable for all disciplines, failure to often audit records, whether or not quantitative or quantitative, can create it difficult for investigators to verify that knowledge collection is proceeding consistent with procedures established within the manual. Additionally, if the structure of communication isn’t clearly delineated within the procedures manual, the transmission of any change in procedures to staff members are often compromised
Quality control additionally identifies the required responses, or ‘actions’ necessary to correct faultily data collection practices and also minimize future occurrences. These actions are less likely to occur if data collection procedures are vaguely written and therefore the necessary steps to attenuate recurrence aren’t implemented through feedback and education (Knatterud, et al, 1998)
Examples of data collection problems that need prompt action include:
• Errors in individual data items
• Systematic errors
• Violation of protocol
• Problems with individual staff or site performance
• Fraud or scientific misconduct
In the social/behavioral sciences where primary data collection involves human subjects, researchers are taught to include one or more secondary measures which will be wont to verify the standard of data being collected from the human subject. For instance, the researcher conducting a survey could be curious about gaining a far better insight into the occurrence of risky behaviors among young adult also because the social conditions that increase the likelihood and frequency of those risky behaviors.
To verify data quality, respondents could be queried about equivalent information but asked at different points of the survey and during a number of various ways. Measures of ‘Social Desirability’ may additionally be wont to get a measure of the honesty of responses. There are two points that require to be raised here, 1) cross-checks within the info collection process and 2) data quality being the maximum amount an observation-level issue because it may be a complete data set issue. Thus, data quality should be addressed for every individual measurement, for every individual observation, and for the whole data set.
