The survey method involves handing out questionnaires to try to get an idea to establish people’s attitude, beliefs and behavior.
You have a population of interest; who you are interested in surveying.
E.g. You want to study Canadian teenagers about their attitude on the legalization of marijuana
Ideally, you would give a survey to every single teenager; however, this is not practical.
What should you do instead? Pick a sample from the population hoping that, with the results that you have gathered from the sample population will be true of the entire teenage population.
You want to generalize your findings from the sample to the population. In order to generalize from a sample to a population, the sample must be representative of the population of interest. This is essential. Your sample must include a wide range of teenagers, varying from their age to where they live and only then, will you be able to generalize your findings.
If you were to only survey a population from Toronto, this would be a bias sample. This data would not be representative of the entire Canadian teenage population because Toronto teenagers might have different attitudes toward the legalization of marijuana than teenagers from Vancouver or Calgary.
How do you ensure that your sample is representative? You do what is called random selection/sampling. You want to randomly pick your sample from the population so that every person in the population has an equal chance of being in the sample.
This is when you want to measure exactly what is going on, and/or describe what you are actually seeing and draw your own conclusions.
- In a naturalistic observation, you can generalize your results. Typically, when you are observing people in a natural setting, subjects are unaware that they are being observed. Therefore, you can reasonably conclude that the way they are behaving in their real-life environment is the way they would naturally behave, whether you are observing them or not.
- You cannot infer cause and effect – you can’t go any further in terms of explaining why things are happening or why people are behaving in a certain way. You cannot say what is causing that behavior because of the fact that you have little control over what is going on.
- You can also misinterpret the situation. How can you prevent this? You can have multiple observers – this is called inter observable/ inter rater reliability. Reliability in this case means consistency.
E.g. 2 kids in a playground and one is head locking the other and you assume that there is aggressive behavior going on but in reality the kids could be talking about what they saw on T.V last night; again, you cannot infer cause and effect.
- It is also time-consuming; you have little control over the situation.
E.g. You want to study bullying in a schoolyard but for days nothing has been happening. You cannot tell a kid to attack another child in order to get a reaction to get some data. You can’t force things to happen if they’re not happening. It is time consuming and this can be frustrating for a researcher.
- The experimenter can tire and loss of attention.
- You can get what is called expectancy effect. This can be problematic with many different methods of research.
E.g. You are conducting a study and trying to prove your hypothesis so you are expecting to see results. Because of your expectations, what you are expecting to happen might influence how you are going to see and interpret your data. Expectations can and do influence your behavior even if you are a researcher.
- Another potential disadvantage of naturalistic observation is subject reactivity. You do not want subjects to be aware that they’re being watched. You do not want the subject to be reacting to your presence because you want them to behave naturally.
In naturalistic observations, sometimes you have what is called a “participant observer”. The researcher who is the observer will act as though they are a participant; in other words they go “under cover”.
E.g. A researcher wants to study how a leader emerges from a group. The researcher will become a part of the social group to be studied to collect information by pretending to be part of the group. However, the researcher must not interfere and/or manipulate the situation to ascertain particular results.