Senin, 02 November 2015
Types of Data in Research based on Scale
Data is an essential ingredient when we conduct a research even though we only employ descriptive analysis. Understanding the types of data therefore is necessary to improve quality of the research. There are many different types of data based on different categories such as based on (1) structure; (2) character; 3) source; (4) collection time; or (5) based measurement scale. This article extracts data based on the scale.
In general, there are four types of data called as NomOrVaRio which stands for Nominal, Ordinal, Interval, and Ratio. NomOrVaRio is not universally accepted but it seems widely used. Explanation for each of those is described as follows.
First, Nominal is the lowest level data because it captures very limited information i.e. IDs or identities or discrete. It differentiates one data from the rest. There is no rank among them as well as no arithmetical equation can be done (addition, subtraction, etc). Usually, we create the Ordinal data by categorization or classification. Statistically, we can count the data as well as find the mode of the data-set. Examples of this data are:
(1) Group of female as 1 and male as 2. We cannot say 2 is better than 1.
(2) Identification for Golkar Party is 1, PDIP Party is 2, etc.
Second, Ordinal has attribute more advance than Nominal because Ordinal data shows a rank (or level) between the data even though it cannot be used in a mathematics expression. Similarly, we create the Ordinal data by categorization or classification. Few examples are:
(1) We set 1 as elementary school, 2 as junior high school, and 3 as senior high school. Therefore we can conclude that 2 is higher than 1.
(2) Customer satisfaction is identified using 1 for very satisfied, 2 for somewhat satisfied, and 3 dissatisfied.
Third, Interval has more characteristics than the previous data. It shows an interval or distance between one data and the others. There is no classification or categorization. Using Interval we can calculate arithmetical equation (addition and subtraction). However, there is no 'true (absolute) zero'. Examples of this type are temperature, IQ scores, test scores. Here is an example.
(1) Test performances are shown as E for 1, D for 2, C for 3, B for 4, and A for 5. therefore we may say that B people has 2 level better than D (4-2=2), but we cannot B is twice as good as D.
Fourth, Ratio is the highest level of data because it has all of characteristics owned by the other data types. It has 'true zero' therefore we can set all mathematics expressions using the data.
(1) Weight of babies A, B, and C are 4 kg,3 kg, and 2 kg respectively. It can be concluded that ratio of baby A and C is 2 (4/2=2).
In general, there are four types of data called as NomOrVaRio which stands for Nominal, Ordinal, Interval, and Ratio. NomOrVaRio is not universally accepted but it seems widely used. Explanation for each of those is described as follows.
First, Nominal is the lowest level data because it captures very limited information i.e. IDs or identities or discrete. It differentiates one data from the rest. There is no rank among them as well as no arithmetical equation can be done (addition, subtraction, etc). Usually, we create the Ordinal data by categorization or classification. Statistically, we can count the data as well as find the mode of the data-set. Examples of this data are:
(1) Group of female as 1 and male as 2. We cannot say 2 is better than 1.
(2) Identification for Golkar Party is 1, PDIP Party is 2, etc.
Second, Ordinal has attribute more advance than Nominal because Ordinal data shows a rank (or level) between the data even though it cannot be used in a mathematics expression. Similarly, we create the Ordinal data by categorization or classification. Few examples are:
(1) We set 1 as elementary school, 2 as junior high school, and 3 as senior high school. Therefore we can conclude that 2 is higher than 1.
(2) Customer satisfaction is identified using 1 for very satisfied, 2 for somewhat satisfied, and 3 dissatisfied.
Third, Interval has more characteristics than the previous data. It shows an interval or distance between one data and the others. There is no classification or categorization. Using Interval we can calculate arithmetical equation (addition and subtraction). However, there is no 'true (absolute) zero'. Examples of this type are temperature, IQ scores, test scores. Here is an example.
(1) Test performances are shown as E for 1, D for 2, C for 3, B for 4, and A for 5. therefore we may say that B people has 2 level better than D (4-2=2), but we cannot B is twice as good as D.
Fourth, Ratio is the highest level of data because it has all of characteristics owned by the other data types. It has 'true zero' therefore we can set all mathematics expressions using the data.
(1) Weight of babies A, B, and C are 4 kg,3 kg, and 2 kg respectively. It can be concluded that ratio of baby A and C is 2 (4/2=2).
