Business Research Method of the of the Reliability and Validity of Measure Reliability de s the constancy of ameasure. An assessment is measured reliable if the same result is yielded recurrently. Reliability can be defined as the proportion of a factual variance to the total variance, which is significant for measurement (McCarthy, 2011, 24). An example to ensure reliability of a measure in a study can be taken into account for further clarification. Suppose a test is designed to measure a characteristic or trait (an example of introspection can be considered).
Each time this test is administered and applied to a subject, the outcomes should be roughly the same. There are four common methods to ensure reliability, each of which evaluates reliability in a diversified manner. They are: 1. Inter-Rater or Inter-Observer Reliability: This method to ensure reliability is used to measure the degree to which different raters or observers provide steady estimations of the similar occurrence. 2. Test-Retest Reliability: This method to ensure reliability is used to evaluate the constancy of a measure for a given timeline or from one occurrence till another.
3. Parallel-Forms Reliability: This method to ensure reliability is used to evaluate the constancy of the outcomes of two tests created in the similar method from the same contented dominion. 4. Internal Consistency Reliability: This method to ensure reliability is used to evaluate the constancy of outcomes across items within a specified test. Validity is the degree to which a test measures what it entitlements to measure (McCarthy, 2011, 24). It is an important property for a test to be valid in order for the outcomes to be precisely applied and interpreted.
Unfortunately, it is impossible to compute reliability precisely, as it cannot be judged on numerical basis but it can be assessed in a number of different ways like other skills are assessed. There are three types of validity to ensure the level of validity in a study: 1. Content validity: If a test is said to have content validity, it means that the items included in the test are representing the complete series of possible things that the test should cover. 2. Criterion-related Validity: If a test is said to have criterion-related validity, it means that the test has verified its efficacy in forecasting measure of the hypothesis.
Concurrent and Predictive Validity are of two kind of criterion-related validity. 3. Construct Validity: If a test is said to have construct validity, it means that it determines a connotation between the test scores and the estimations of a theoretic characteristics. Advantages and Disadvantages of Stratified and Random Sampling Sampling approaches assist reports to be made for a collection or group, on the basis of data collected for a definite percentage.
These outcomes are applied to the non-destructive testing of arrangements for fault and error detection in the specified samples. Stratified and random, both are types of sampling that offer numerous benefits and disadvantages. Stratified sample can deliver better precision and involves smaller sample that saves money. It can safeguard beside an unreliable sample, for example extracting all female sample from a mixed-gender populace (Brown, 2009, 332). The major disadvantage that stratified sampling have is that it may involve extra administrative determinations than other sampling method demands. There is significant reason for favoring the random sample since it imitates to the theoretical representations.
The random sampling is free from the prejudice that may be presented by inaccurate space weights, and inclines to signify in all of its many features. A single random sample can be used for many different purposes with mathematically determinate limits of error. There are number of advantages and disadvantages of random sampling, and certainly, many types of random sampling itself as t samples the whole populace. The advantage of this method is that if all those asked offer a sample then the outcomes delivered will be extremely demonstrative.
The disadvantages of this type of sample is that it is uneconomic and not efficient to be accomplished, and also that the period to get the outcomes would be too extended. Descriptive Analysis Descriptive analysis is an approach to quantify and identify attributes of a product with the help of panelists precisely skilled for the study. It can involve analysis of all parameters related to the subjects and can also be limit to definite aspects. These parameters can be smell, taste, texture, thickness, state or any feature that a subject can depict.
As descriptive analysis is a sensory approach, many descriptive methods are presently engaged by sensory specialists. All descriptive analysis methods include the finding and portrayal of both the qualitative and quantitative sensory characteristics by a qualified panel of juries. An example of wine descriptive analysis may consider for additional clarification. The descriptive analysis of a wine may be taken to quantify and identify its attributes such as its look, aroma, taste and texture properties which distinguish it from other alcoholic drinks.
In addition, panelists segregate and rate the intensity features of a wine sample and define to what degree each property of that sample scores. Multiple samples of different wines may hold the similar qualitative descriptors, but they may vary evidently in the concentration of each, thus causing in moderately unalike and distinguishing sensory profiles of each wine. Cross-tabulation report is required to display the relationship between two or more studies. It is used to analyze survey questions of two or more researches and extracts new relationships from these surveys.
The important aspect of to a cross tabulation analysis is to identify the purpose of the study. With the help of cross tab analysis, entire survey can be successfully analyzed by selecting the appropriate part of the audience. Statistical Analysis Statistical analysis is used to analyze data with one or multiple number of variables to acquire a desired result and output (Jones et al, 2006, 867). Univariate analysis is a statistical analysis method used to analyze data with one single variable during the reserach. In research studies, there is a need to only isnvolve one variable at a time.
For example, if a researcher wants to report the total number of suicides in 2010, he will only have a single variable; suicides in the year 2010 that will be used in the reports. This type of statistical analysis is known as Univariate analysis. Bivariate analysis is a statistical analysis method used to analyze data with two variables at a time. This approach discovers the perception of association and dependencies between two variables. Dependencies are based on the way these variables instantaneously change together.
This association is known as co-variation. For example, if a researcher wants to report the total number of infant mortality in 2010, he can define two variables for a clear picture, one for male infants and other for female infants. This type of statistical analysis is known as Bivariate analysis. Multivariate analysis is a statistical analysis method used to analyze data with more than two variables at a time. This method is used when the researchers needs to provide a deep study with multiple factors affecting the purpose.
For example, to analyze the properties of wine, researchers need to examine its look, aroma, taste, post taste and texture etc. This type of statistical analysis provides details regarding each of the variables defined and specifies the relationship between all the defined variables. References Brown, G. (2009). The journal of Marketing: A Comparison of Sampling Methods. University of Chicago, 332-336. Jones, M., Onslow, M., Packman, A. & Gebski, V. (2006). Guidelines for Statistical Analysis of Percentage of Syllables Stuttered Data, Journal of Speech, Language, and Hearing Research, (4): 867-78 McCarthy, L.
(2011). Clinical Practice; Research from Princess Alexandra Hospital Broadens Understanding of Clinical Practice. Health & Medicine Week (12): 24-27.