Guide to conducting your own study Part 2
Making a hypothesis: Once you have decided on a question, done some research, and decided how to operationalize your variables, you are now ready to make your hypothesis. If you have successfully completed the previous steps, this process should be pretty painless. First, if you are studying two variables, you must determine which is the dependent and which is the independent variable. If you figure this out, you are already half way to a hypothesis. Your hypothesis so far is: “y” depends on “x”. Now, you must look to your research and take an educated guess on how “x” affects “y”. As “x” values increase, what do you suppose will happen with “y” values? Will “y” values increase or decrease? You hypothesis should now be: “y” depends on “x” such that as “x” values increase/decrease, “y” values will increase/decrease. For example, in my guinea pig experiment, I might say, “over 30 days of feeding, guinea pigs that received more food had a higher weight than those receiving less food.” That is, as the “x” value amount of food, increaces, so does the “y” valueweight. Finally, you must operationalize your hypothesis first by stating by making a specific prediction in the terms of your variables. For example, in my guinea pig experiment, I would say “in order to support my hypothesis, the guinea pigs the received 10 grams of food each day should weigh on average at least 20% more than the guinea pigs that received 5 grams of food.”
Pick your sample: As much as you might like to collect data on the entire population that you want to study, this will certainly not be a possibility. You will therefore have to pick a sample of your population to study. The way you pick this sample and the size of this sample will affect the reliability of your data and how well you can generalize the results of your sample to your population. There are a number of different types of sampling that fall into two categories, probability sampling and nonprobability sampling. Probability sampling includes: Random sampling In a random sample all members of the population you choose to study have an equal chance of being selected There is no bias involved in the selection of the sample Any difference between the characteristics of the sample and those of the population happens strictly by chance stratified sampling Stratified samples attempt to create a sample that is purposefully representative of the population The researcher determines what characteristics are important to their study and selects randomly from different groups accordingly Nonprobability sampling quota sampling a researcher wants to be sure to include people of certain characteristics in their population so they set a quota (a set amount) of people of a certain characteristic to include in the sample purposive sampling (including snowball sampling) a nonrepresentative subset of a population sample constructed with a specific purpose in mind convenience sampling a convenience sample is just that, one based on convenience also called an accidental sample not representative of the population Depending on what population you decide to study, it may or may not be possible for you to pick a representative sample, and for the purposes of this project, that is totally fine. What is most important with regard to sampling is that you report the shortcomings of the sampling technique you use in the final presentation of your data. Check out this link for more info on sampling: http://psychology.ucdavis.edu/sommerb/sommerdemo/sampling/types.htm
Design your study Once you have picked your question and operationalized your variables (decided how you are going to measure them), you are now ready to design your study. You must determine what is the best method for gathering data. There are many possible methods for collecting data. Data collection falls into two broad categories, qualitative and quantitative. Qualitative studies focus the reasons for and motivation behind different phenomena. Their results are not usually able to be generalized, but they can be very helpful to gain better understanding of things we know little about Quantitative studies focus on measurement and data collection. They can be executed precisely and with minimal systematic errors, leading to an ability to generalize results from a sample to a greater population. The main drawback to quantitative research can arise from the need to isolate variables. When we only focus on the relationship between two things (for example the food intake of a guinea pig and its weight), we can miss the complex interactions between variables that we see in the real world. Data collection methods include, but are certainly not limited to: Surveys and questionnaires Interviews Focus groups systematic observation tests experiments As with sampling, there is not one “best” method for collecting data, but rather different strengths and limitations for the different methods. Most students will use a survey to gather data, as this is most convenient and practical for your projects. If you decide to create a survey, make sure to refer to the packet you received in class or one of the following sources for help:  http://www.mathsisfun.com/data/surveyquestionnaire.html http://www.cc.gatech.edu/classes/cs6751_97_winter/Topics/questdesign/ Make sure you have a clear procedure for how you will administer your survey/experiment and ho you will record your data.
Complete your study and collect your data Once you have designed a study to collect your data and picked your sample group, it is time to gather data. This should be very straightforward, as you know what data you will be collecting and who you are collecting it from. Make sure you have a clear method of keeping track of your data as you gather it. Make sure you precisely follow the procedures that you laid out for picking your sample, conducting your research and recording your data.
