Research is important in every area of the society. It enables people and governments to solve problems appropriately by choosing the most appropriate solution to any problem affecting the population. It also helps in good management of resources. Research involves obtaining statistical values and analyzing them.
This is done by first identifying the research that one wants to conduct; this includes an introduction of the study with a problem statement, hypothesis, definition of terms and a summary. The introduction is supposed to give the reader the background information and objectives of the research.
A section of the method to be used in the research should be included. This include the design, subject of participants, instruments and the procedure showing the plan and the activities which will be involved in the study. This should be followed by some idea about the data that will be collected. A discussion and limitations of the research study should also be indicated (Salkind, 2010).
Literature review is important as it reveals that the researcher understands the research problem (Salkind, 2010). Identifying the appropriate sampling population and method for data collection is significant. It may be interviews with questionnaires. The data is then collected and finally analysis of this data carried out.
A research manuscript is written after the research is done, data analyzed and conclusion and recommendations done. At this stage it is ready to be published. The research manuscript is written by identifying a title which describes the research concisely, clear and attractive to the readers.
The researcher should also give a summary of the research precisely. It should also include an introduction, materials and methods. Results, discussion, references and appendices must be included. Author notes, footnotes, table captions, tables, figure captions and figures are also important to enhance understanding of the research carried out (Salkind, 2010).
Data collection and analysis is the base of statistics. Statistics therefore include collection, organization and interpretation of the data collected. There are two types of statistics namely descriptive and inferential statistics. Inferential statistics are used to make conclusion in the results of data whereas descriptive statistics describe the qualities of a given sample (McMillan, 2005).
Effective research ensures that the collected data represents the population on which the research is being done (Salkind, 2010). For example, if one wants to conduct a research in teenagers in a particular district it is important that there is equal distribution in sampling areas including both boys and girls. From the data that one collects, one is able to make a conclusion if for instance it was drug abuse the percentage of teenager abusing drugs in that population.
Data interpretation should always avoid assumption, but to some extent one may conclude that the result or probability that a person is a teenage is a drug abuser is a matter of chance, this occurs when one is not able to find any evidence that it is because, e.g. of family background that the teenage is not abusing drugs.
A research always works towards avoiding any assumptions by ensuring that the questionnaires cover many issues related to that research to improve the accuracy. The researcher should also try to collect his data from a normal population.
Once sample data has been gathered through an observational study or experiment, statistical inference allows analysts to assess evidence in favor or some claim about the population from which the sample has been drawn. The methods of inference used to support or reject claims based on sample data are known as tests of significance. (Test of significance, n.d., p. 1)
This test of significance is subject to null hypothesis H0, which is believed to be true but has not been proved. Since sampling is not 100% perfect and does not cover every sample that is available, one is likely to make some errors. The null hypothesis is subject to acceptance or rejection. If the null hypothesis is true and from the results is rejected, this is known as type I error. It has some level of significance, which are conventional values associated with it. When one accepts a force null hypothesis, this is known as Type II error (Salkind, 2010)
Descriptive research is a kind of research that works to explain a give research in relation to other environmental factor. However it does not restricted to any group such as teenagers .For example the use of alcohol and drug abuse in a given area regardless of age or gender (Salkind, 2010)
Survey research on the other hand studies intensively. It even involves the thoughts of the respondent while conducting a survey research. This calls for an intensive interview to the respondents and needs to be handled with a lot of care and attention. The interview questions are closed or open ended.
To increase efficiency and effectiveness, there is need for clarity of objectives and an appropriate sample is necessary. In addition one needs to come up with questions that are simplified or structured but should be easy to analyze and directed to particular goals (Salkind, 2010).
Correlation research is the research that measures the degree of linear relationship between two variables. The correlation coefficient varies from -1 to 1.The positive correlation is shows that there is a positive relationship between two variables while negative correlation indicates a negative association between two variables (Salkind, 2010).
To be effective in research, one must find an appropriate research method. These experimental designs include pre-experimental, true experimental and quasi experimental which is also known as casual comparative design (Salkind, 2010).
Pre experimental design is not characterized by random selection of participants from a population and does not involve a control group. True experimental designs include all the steps in selecting and assigning subjects in a random fashion, plus a control group leading to a greater efficiency and comparison (Salkind, 2010).
Quasi experimental method is different from pre-experimental and experimental methods in that the hypothesized cause of differences you might observe between groups have already occurred. This means that pre assignment of groups has already occurred. There are different designs that can be used in quasi experiments.
The most common is non equivalent control group design especially when it is impossible or difficult to assign subjects randomly to groups. Static group comparison design is used when one cannot randomize and cannot administer a pretest (Salkind, 2010).
Single subject research design is a good method of understanding causal relationships that looks at individual rather than groups. They are mostly used in behavioral analysis and education. Multiple baseline design is a good design where two behaviors, two subjects or two occasions are selected for a study and a treatment applied to one of them. In this way the behavior, the participant, or situations in which treatment is not present serves as a baseline against which the effects of treatment can be determined (Salkind, 2010).
In developmental research one may apply longitudinal method. This determines changes in behavior in a group of subject for more than one point in time. A cross sectional method can also be used. This is done by examining one group of people repeatedly over a time. A good combination of the methods in a particular study enhances accuracy of the results.
For one to obtain accurate results, it is important that to consider the, most appropriate method to carry out the research. Inferential statistics help one to make the right decision. The research project should therefore describe its purpose clearly on the methods and activities so as to accomplish its purpose.
McMillan, J. (2005). Data Analysis and Collection. Retrieved on May 14, 2011 from http://staff.ed.uiuc.edu/jmmcmill/portal/datacollection.html
Salkind, N.J. (2010). Exploring Research. New York, NY: Prentice Hall.
Tests of Significance. (n.d.). Tests of Significance. Retrieved on May 14, 2011 from http://www.stat.yale.edu/Courses/1997-98/101/sigtest.htm