Likert scales are psychological tests and instruments of measuring that are commonly used in measuring attitudes. The attitude expressed is by means of opinions by customers. What really is paramount is the attitude of who believes. The scale measuring attitudes analyzes the feelings, thoughts and opinions of the person to the facts already specified. The scale measures individual attitudes or predispositions in particular social contexts or events.

The main limitation of data collected through likert scale is generally grouping data collected into different hierarchies mostly levels of five point scale. The five point scale has usually been shown to reach the upper limits. The standard and mean deviation are parameters for descriptive statistics which are invalid (Echon, n.d.). If the number of people interviewed is small then there will be a huge variation in the score. The other limitation is many people will opt not to respond to the question due to the processes involved in logging in the websites to respond to the questions. Many will also opt not to respond to the questions ignoring them.

Although the staff members may appear very knowledgeable in regards to the products offered by the store, the customers are less aware regarding the staff level of knowledge for the products. One hundred customers were interviewed concerning this question and the results was an overwhelming response that they agree to the fact that employees are not aware of the product.  Io of the customers strongly disagrees, 25 of the customers at 25% disagree, 3 customers are unsure, 17 at 17% agree while 5 of the customers strongly agree. When the data is compiled into a single score, the score is calculated by multiplying and adding the five responses together and the dividing the total results by the number of respondents which is 100 customers. The single score received is 3.19. The reason why this calculation may be incorrect is that the number of customers interviewed is very small. 100 customers is not a true representation of the entire population of customers. This number is too small to find the correct single score and therefore the single score earlier reported. An increase in the number of customers responding will change the single score to be less than 3.19. For example, if the number of customers increases to 200, then the single score will be (319/200) which is equivalent to 1.595.

I strongly disagree with this likert scale and the scre given of 3.19. this is due to the small number of people in the survey. The number will not represent the real number of customers.

Although a survey takes a sample of the population in question in this case the customers, the higher the number of customers took for the survey, the higher the chances for an accurate answer. 100 customers are very few to get the correct impression in this scenario.  Through the likert scale, the number of people making all the contribution should be high to get the true representation of what kind of response the customer feel toward the knowledge of employee knowledge  of the products (Echon, n.d.).

The reason why this number will give an inaccurate representation is because it is a very small number compared to the total customer base. Secondly, voting online or through the company’s website does not show the cluster or demographics of the population. For example, more males may vote more than the females while in actual sense the females shop more than the male customers and would therefore make a more informed decision. Additionally, interactions between the customers and the employees are of different levels. The law of attraction states that unlike poles attract, therefore, many male customers may interact more with the female employees ignoring their male colleagues.  The resulting survey may be skewed or biased, more so, with the small number of customers in the survey.

The data is not very reliable while the likert scale is not as conclusive as it should be to give a concrete answers due to its nature of varying responses. The data is not very reliable to be used due to the small number of correspondence.  A survey should have the maximum number of people to represent the entire population of the area of survey. The percentages being used in the data are not accurate while the customers who vary in agreeing and disagreeing are considerably very low. Therefore for reliability of the data, the number of customers in the survey should be increased while the likert scale should have a dominant answer, for example, disagree at 70%.

We discussed four types of quantitative data in class – nominal, ordinal, ratio and interval. What types of data are collected from each of the following questions? Justify your answer.

A nominal variable is one that “distinguishes between ordering them subject to a limited number of categories, including the type or class.” When a nominal variable is an immutable characteristic trait or a participant research, the term “variable attribute” is also applicable.

As an example, a data worktable with results of the tests in a column. The other 2 columns show the gender of every evaluation taker as well as the kind of test run. The outcomes of the studies are not nominal data since they provide particular figures that may be added in, normal or even put into numerical progression. Gender is a nominal variable, since it basically categorizes evaluation takers in a set selection of groupings: male and female.

Conversely, quantitative variables calculated rather than explaining the data. The interval parameters are quantitative parameters missing a practical 0. The temperatures scales are an outstanding example, in which the Fahrenheit as well as Celsius scales are not the same and also arbitrary 0 points. Lastly, the proportion parameters are quantitative variables intrinsic with a 0 point showing the lack of the presented character. Fahrenheit falls short of a practical zero thus; they could be categorized just as interval data. An ordinal variable, for example a studying level determined by the qualification, places the data in the perfect order, however the distance is not quantitative ranges. For instance, the reading level class eight is possibly not twice as tough to grade four. 

A ratio variable, has got most of the attributes of an interval variable, as well as possesses a definite description of zero .0. While the variable is the same as 0 .0, there is certainly nothing of this variable. Parameters such as height, heaviness, enzyme action are ratio variables. Temperatures, stated in F or even C, are not a ratio variable. Temperatures of zero .0 on both of these scales does not necessarily mean ‘no heat’ Nevertheless , temperature in Kelvin is a ratio variable , since zero .0 Kelvin does indeed indicate ‘no heat’ . Ratio: This is data range with a natural zero point. For example, time to find a product on a website is the relationship, because time is 0 significant. Kelvin has a point 0 (absolute zero). The steps in both these scales have the same degree of magnitude

 this is categorized as Ordinal variables: designated categories, but have the additional property of allowing the sort categories from the largest to the smallest, from best to worst or the first to the last (Cramer, 2003). Ordinal variables considered common classification of social class (high, medium, low, indigent), level of education class (last year, first year, etc.) and housing quality (standard, insufficient in ruins)

Bank account balances is an interval variable.The interval variables give a sense of “how much” or “that size” which so hot, so obstinate, so depressed, that so long and so heavy. With the interval variables you think in terms of distance between scores on a straight line. A good example for this is the bank balances.With the interval variables you can add, subtract, multiply and divide and calculate averages scores, which is not possible with ordinal variables.

Both the interval variables have reason as set intervals in a measuring unit; but only reason variables include meaningful zero point.Some interval variables can have a zero score, but the zero point is arbitrary; that is, it could be placed anywhere within the possible range of a variable because zero does not mean “no”. True zero points reason variables allow greater flexibility in the calculations and statistical analysis (Byrne, 2002). As the interval variables, variables can multiply and divide reason, but can also be calculated reasons.

Assume you are the coach of a local sports team. You believe it is possible that drinking orange juice three times per day, for four days each week, might make the players perform better in the game at the weekend. Explain how you could test this hypothesis in each of the following ways: a. As a descriptive non-experimental study b. A quasi experimental study c. An experimental study Which of the studies would give you the firmest evidence of whether your theory is correct? Justify your answers 

A descriptive research is the one where the details are gathered without swapping the environment ( that is , hardly any manipulation ) . Commonly known as like “correlational”or even researches “observation .” It is usually a descriptive research since “any research which is not really experimental .” In our study , a descriptive research might givedetails about common wellness status , behaviour , actions or even additional features of a particular group . Descriptive studies are likewise performed to establish associations or connections between issues in the environment . In this case a non descriptive study is the most effective .

Therefore, it is a type of research that shares many of the characteristics of an experiment, but the comparisons in the response of subjects are made between groups ‘non-equivalent’, ie groups that can differentiate into many other aspects in addition to the ‘exposure (Allen, Titsworth, & Hunt, 2009)’. The main difficulty will be to distinguish the specific effects of treatment ( ‘exposure’) of those specific effects that result from the lack of comparability of groups at baseline and during the study, which compromises the internal validity of the study. In case there is no control group, it can not ensure that the changes appeared to be due to the intervention itself, or to other interventions or uncontrolled factors. It is not effective in this case.

In experimental studies manipulation of a particular exposure in a group of individuals is compared with another group that was not operated, or that is exposed to other intervention occurs. This type of quasi-experimental designs are usually (but not handling randomization exists), in which one or more groups  will receive the intervention, while others serve as control. Experimental studies have a carefully designed sufficient sample size, a process suitable randomization, intervention and perfectly controlled monitoring can provide very strong evidence that allow us to study issue judgments about the causal relationships between variables.The validity of this study lies primarily in the random process that make comparable groups  the most important variables in relation to the problem to study. Therefore, this type of study is not suitable for this type of study because there are no comparable groups and sample is not large.

Allen, M., Titsworth, S., & Hunt, S. (2009). Quantitative research in communication. London: SAGE.

Byrne, D. (2002). Interpreting quantitative data. London: SAGE.

Cramer, D. (2003). Advanced quantitative data analysis. Maidenhead, Berkshire, England: Open University Press.

Csatar, P. (2014). Data Structure in Cognitive Metaphor Research. Frankfurt: Peter Lang GmbH, Internationaler Verlag der Wissenschaften.

Echon, R. Advances in image analysis research.

Quantitative research. (2010). [Place of publication not identified].