Data Collection
The data
used in this study was collected through a survey which I conducted according to
Donald C. Orlich's book Designing Sensible Surveys (1978). In this
book, surveys and questionnaires are discussed, including the advantages and
disadvantages of using surveys, avoiding bias, question placement, and
statistical analysis.
I conducted this survey because the data I needed
is not available at another source. The objective of my survey is to
determine the location, age, and opinions (both general and specific) of the
Cuyahoga River of people of all ages, male and female. I wanted to know if
the amount of time one has spent observing the Cuyahoga River affects that
person's opinion of the River. The two options I had of ways to collect
this data were either through surveys or through interviews. I decided to
use a survey because of the four following factors: each respondent
receives identical questions, uniform data are collected, responses can be
easily tabulated, and a survey is less expensive and less time-consuming
(for respondents and data collector) than an interview technique.
I designed the survey to be simple and benign, so that there
would not be any confusion or apprehension for the respondent.
According to Orlich, "Questionnaires are not designed to instruct the
respondents about issues, techniques or problems. They are designed to
gather opinions or attitudes about already known issues, problems, or events" (Orlich, p. 25). Once I had my survey designed, I
showed it to my advisors and to relatives and friends to help find any unclarity
or bias within the survey. Once I had a final version, I made a large box
with a tablet of surveys, pencils, and a slot for completed surveys. This
made it easy for people to not feel threatened, pressured, or
hurried. I then took the box and additional survey tablets to
locations where I collected the data. All data collection was in
accordance with regulations set by the Kent State University Human Subjects
Review Board.
On the survey, for the questions of opinions, I used Likert
scales. When I entered these opinions into my database, I assigned a
numeric scale to the Likert responses, for ease in tabulation and
calculation. According to Orlich, "Likert scales and ranked data are
frequently summarized by determining the average mean scores as a measure of the
central tendency" (Orlich, p. 139). This is how I was
able to manipulate the data, calculate averages, and use statistical analysis
for the opinions gathered.