Data collection

Data collection is the process of obtaining and analyzing information on specific variables in a predetermined, methodical way so that one may subsequently analyze results and respond to pertinent queries. Data are the gathered pieces of information. Two methods can be used to gather data:

• Subjective data collection: Subjective data are those that are gathered from patients and their families. It comprises their perceptions of worries, sensations, and thoughts.
• Objective data collection: Data is gathered objectively through records, observations, interviews, reports, etc.

Sources of data

Primary data:

Primary data are the specifics that are entirely and directly relevant to the issue at hand. Primary data are considered to be authentic information and are used as the fundamental input for analyzing and resolving any issue pertaining to marketing-related operations. Data are gathered by:

• Observation
• Interview

Functions and Characteristics of primary data:

• Original data are primary data.
• The cost of primary data is high.
• Primary data collection requires a lot of time and effort.
• They are gathered in light of the current issue.
• They are purposefully gathered from pertinent responders.
• Primary data are gathered using certain techniques (such as surveys, observations, experiments, etc.) and equipment (such as printed forms, questionnaires, cameras, etc.).
• They serve as a fundamental input for the study.
• Before being used, they must be provided, processed, or evaluated.

Secondary data:

On the other hand, published data are considered secondary data. They are simple to locate. They don't need to be processed or analysed before usage. Instead of being generated, they are collected. Secondary data are specifics that have been gathered for purposes other than a particular research issue. The recorded data is another name for them. They have been made public. They have been gathered by others for their current issues. They may not always be helpful. They are additional information to primary data. They back up the raw data. Data are gathered by:

Census:

• Vital registration of events
• Journals
• Hospital records
• Reports

Functions and Characteristics of Secondary Data:

• For the research at hand, secondary data are published data rather than original data.
• They provide the most recent data.
• They are conveniently obtainable through a variety of internal and external sources.
• They need less work, time, and money and are comparably inexpensive.
• They had previously been gathered by other people for their own issues and circumstances.
• They provide as additional main data. They are mostly employed for problem definition and problem understanding.
• Secondary data usage is optional. Even without the usage of this type of data, research can be done.
• They need to be analyzed before use; they can be utilized without processing. They are immediately usable.
• The three key issues are timing, accuracy, and relevance.

Functions and Characteristics of Qualitative and Quantitative Data

Anything that can be quantified or expressed as a number is considered quantitative data. Scores on aptitude exams, study hours, and subject weight are a few examples of quantitative data. These data are amenable to the majority of statistical manipulation and can be represented by ordinal, interval, or ratio scales.

Numbers cannot adequately express qualitative facts. Typically, qualitative data are those that indicate nominal scales, such as gender, socioeconomic level, and religious preference.

Both methods of measurement are employed in academic publications for education. Only quantitative data may be statistically evaluated, allowing for additional evaluations of the data.

Census and sample survey

A census is a study of every component, be it individual or collective, within a population. It is referred to as a comprehensive enumeration, which is another word for a full count. A sample is a subset of units chosen to represent all the individuals in the population.

A sample survey is one that is conducted using a sampling technique, meaning that only a section of the population is surveyed.

Methods of sampling

• Probability Sampling

Every unit in the population has a possibility (higher than zero) of being chosen for the sample in a probability sample. Randomly choosing components from a population is the basis of probability sampling. An equal and independent chance of getting chosen exists for each member of the population in a random selection process. It improves the sample's representativeness.

Popular probability sampling techniques include the following:

Simple Random Sampling:

Simple random sampling is the process of selecting a sample of size n from a population of size N in which every feasible sample of size n has an equal probability of being chosen. A simple random sample is what is obtained in this way. The majority of sample designs are based on simple random sampling. These two techniques can be used to accomplish simple random sampling:

• Simple Random Sampling Without replacement (SRSWOR)
• Simple Random sampling with replacement (SBSWR)

Stratified Random Sampling:

The population under investigation is not uniform very often. In that case, we take into account several parts or features that are homogenous within themselves to make the sample more representative. The stratified random sampling method is applied in these circumstances. The basic goal of sampling theory is to get information with the least amount of money, time, and effort. Therefore, the best way to improve precision is to decrease population variability, which can be accomplished by stratifying the population into different strata. To stratify is to divide into layers. When three criteria are satisfied, a stratified sampling strategy is most successful:

The variables which the population is stratified are strongly correlated with the desired dependent variables.

Variability between strata maximized:

Continual Sampling A 1-ln k systematic sample with a random start is one that was created by randomly choosing one element from the first k elements in the frame and every kth element after that. If a complete and up-to-date list of the sample units is provided, systematic sampling is a frequently used strategy. This entails choosing only the first unit at random, with the remaining units being automatically chosen according to a predefined pattern with uniform spacing between units. In general, systematic sampling includes choosing one element at random from the first k elements and then choosing an element every kth after that. Simple randor sampling is a poor substitute for systematic sampling for the following reasons:

• Systematic sampling is easier to perform in the field and
• Systematic sampling can provide greater information per unit cost than simple random sampling can provide.
• It has less selection errors by field workers than are either simple random sampling or stratified random samples, especially if a good frame is not available

Cluster Sampling:

The purpose of a sample survey is to gather a certain amount of data on a population, with a minimum requirement. For the three reasons outlined in the previous section, stratified random sampling is frequently more appropriate for this than plain random sample. More information is available per unit cost with cluster sampling than with any of the other three previously discussed designs. The sampling process may be very challenging and time-consuming in large-scale studies where the population is dispersed geographically. Obtaining a complete listing of some populations may also be challenging or impossible. Consider a researcher who wanted to speak with 100 nurses who worked in Nepali hospitals. If 100 names were chosen through a simple random sampling procedure, it is quite likely that the investigator would face with travelling several hospitals in 75 districts of Nepal. This would be very expensive and time consuming activity.

Multistage Sampling

So far, we have only looked at sampling techniques that count every component of the chosen clusters. The cluster sampling approach can be developed in a variety of ways to meet the demands of circumstances that are more complex. Multistage sampling, a sophisticated type of cluster sampling, involves the nesting of two or more tiers of units inside one another. We found that the efficiency of the cluster as a sampling unit typically decreases with cluster size. As a result, it makes sense to assume that for a given set of elements, better precision will be reached by selecting a small number of clusters and then sampling a large number of elements from each of them.

Probability Proportional to Size (PPS) Sampling

For a basic random sample, each unit in the population has an equal chance of being chosen. However, simple random sampling does not account for the potential significance of the larger units in the population if the units vary significantly in size, which is frequently the case. There are many ways to use this auxiliary data to get estimates of population parameters that are more accurate, that is, have fewer standard errors. Giving the various units in the population varying selection probabilities is one such technique. Probabilities of selection may be assigned proportionally to unit size when units vary in size and the variable under research is prima-facie associated. Such a procedure in which the units are selected with probabilities proportional to some measure of th size is known as sampling with probabilities proportional to size (PPS). 

Methods of  Data Collection:

• Use of questionnaire
  • The primary tool used in survey research for data collection is the questionnaire. Essentially, it is a collection of standardized questions, also known as items, that adhere to a predetermined format in order to gather individual data regarding one or more particular themes. Interviews and questionnaires are sometimes used interchangeably.
• Observation with checklist
• An observation checklist is a list of things that an observer is going to look at when observing in the method of data collection.
• Interview: 
  • Focus group discussion
• Participatory Rural Appraisal (PRA):
  • The PRA method puts an emphasis on local knowledge and empowers locals to do their own assessments, analyses, and planning. It encourages information exchange, analysis, and action among all stakeholders by using group exercises and animation.
  • Steps of PRA:
  • Physical arrangements:
    • It takes into account nearby lodging, plans for a meal, transportation, portable computers, and funds for refueling and purchasing materials like flip charts, paper, and markers.
  • Trainings:
    • Trainings simply refer to reviewing prior information, putting it to use, and passing it on to others. The participating PRA team needs to receive training for information gathering as well as other tasks.
  • Workshops:
    • A brief workshop on the introduction, goal, and activities of the field work should be held by PRA.
  • Data collection and analysis:
    • It is necessary to gather, analyze, and interpret information. Information should be gathered utilizing a variety of methods and technologies.
  • Taking notes and reporting:
    • After fieldwork, a preliminary report is written, and a thorough report is then created. Within a week or so of the fieldwork, a preliminary report should be available, and the final report should be shared with all participants and the local institutions that were involved.
  • Rapid Rural Appraisal (RRA):
    • RRA is a collection of methods that can be used as the first phase of surveying. The methods mainly entail a fast, informal exploration of a given geographic area. As a result, the study develops a "knowledge" of the local circumstances, issues, and characteristics of the rural community. They can offer fundamental details on whether starting a survey project in a particular location is feasible.