Levels of Measurement

          Measurement is the use of numbers to represent things. These "things" are usually referred to as concepts. In sociology, there are three main levels of measurement which include nominal, ordinal, and interval. One main aspect of understanding the three levels of measurement starts with knowing that if a variable isn't measurable, it isn't a variable. Variables are the actual ways we measure concepts. Levels of measurement also go hand-in-hand with measures of central tendency. Each level of measurement correlates with either mean, median, or mode.

          

          The nominal level of measurement is measurement in its very simplest form. Basically you rename a category with a number. For example, the division of precincts in the state of Louisiana. Northwest Louisiana could be labeled as a 1, northeast Louisiana could be labeled as a 2, southwest Louisiana could be labeled as a 3, southeast Louisiana could be labeled as a 4, and central Louisiana could be labeled as a 5. These numbers means nothing more than a categorization for different precincts in Louisiana, therefore, central Louisiana being a 5 and northwest Louisiana being a 1 does not mean that central Louisiana is better because 5 is greater than 1. You could even label the numbers as 34, 87, 192, and 3 and there would be no difference in meaning.

          The ordinal level of measurement is very similar to nominal in the case that you can rename a category with a number, however the number actually means something in the ordinal level. For example, you could be given a survey that asks, "What is your income per year?" The answers could be coded as 1 meaning $10,000-$29,999, 2 meaning $30,000-$49,999, 3 meaning $50,000-$69,000, and 4 meaning $70,000-$89,000. The numbers in this level of measurement mean quite a bit because they are on a scale. The person getting surveyed can assume that a "1" is less than a "2" and a "2" is less than a "3" and a "3" is less than a "4" and vise versa. Because of this, it is critical that the numbers are in order from least to greatest. The numbers would be labeled as 21, 34, 56, and 67, but as long as the numbers are still in order from least to greatest, the scale would still be valid. 
         The interval level of measurement is the most advanced of the three main levels. This level includes everything from the first two levels, however, the difference of amount between two particular numbers are viable and countable. For example, IQ test scores. A score above a 145 is genius and a score below a 70 is borderline deficiency. If someone made a 135 and another person made an 80, you could easily tell that the person that made a 135 was obviously much more intelligent than then person who made an 80 due to the fact that the difference in these numbers and viable and countable. Knowing a range is also extremely important in interval level of measurement. If someone told you scored a 135 on their IQ test, that number would mean nothing to you unless you know that an IQ test ranges from 145 to 70. 
          Ratio level of measurement is very similar to the interval level, but it is not one of the three main levels of measurement. Ratio adds the concept of an absolute zero. This means that you can construct a meaningful fraction with a ratio variable. For example, one inch equals one mile on a map. This ratio of one inch equals one mile shows the concept of an absolute zero because it gives an exact and equal measurement to a mile, just in smaller terms.

         

          The three ways of calculating averages include mean, median, and mode. Mean is the value obtained by dividing the sum of several quantities divided by the number of quantities. Median is the midpoint of a frequency distribution of observed values or quantities after putting them in numerical order. Mode is the value that occurs most frequently in a given set of data. Mean is the average to use when working with interval and ratio levels of measurement, median is the best when working with ordinal level of measurement, and mode is the best when working with nominal level of measurement.

          Mode is the best average to use when working with nominal level because it determines what numerical value occurs most commonly. For the example for nominal level, I divided Louisiana into five different voting precincts. If you wanted to see which precincts casted the most votes, mode would be the best average because it would show you which precinct voted most commonly. 

          Median is the best average to use when working with ordinal level because it determines a midpoint. For the example of ordinal level, I used incomes. Median is very important when working with this because if you went to a job interview and asked what the average salary is, you would want them to tell you the midpoint of the salaries as opposed to the most common salary or the mean of the salaries. This is because if there are six workers at a particular business and the salaries of the six workers are $300,000, $80,000, $55,000, $40,000, $20,000, and $20,000, the mean would be $85,833 and the mode would be $20,000. However, the median would be $47,500 which is obviously the most accurate of the three. 

          Mean is the best average to use when working with interval level because for example, IQ scores. If you are trying to find the average IQ score of six different people which include the scores of 140, 140, 140, 100, 90, and 80, the mode would be 140 and the median would be 120. However, the mean would be be 102 which is definitely the most accurate for this set of data. 

         Not only is choosing the correct level of measurement important when working with a set of data, but choosing the right measure of central tendency to use on a level of measurement is extremely important as well. As shown in these examples, data and averages can be skewed if the correct measures are not used. I, personally, use levels of measurement and measures of central tendency quite often without even realizing it. For example, my microbiology class has four grades, and I had taken three of them, but I wanted to find out what grade I could make on the fourth test to still make a B. The first test I made an 82, the second test I made an 80, and the third test I made a 78. I realized that this set of numbers should be used in the nominal level of measurement, therefore, mean would be the best measure of central tendency. I found the mean of the numbers and realized that I would have to make an 80 on the fourth test to still make a B in the class. This shows just how important levels of measurement and measures of central tendency can be. This also shows that these terms are not just used by sociologists, but by normal people in their everyday lives as well.