We love levels of measurement
This is the level of measurement of this variable: Your respondent's favorite outdoor recreation activity.
Nominal.
I wrote a paper about hurricanes and sexism, but I manipulated my logistical regression model so my results fit what I am arguing. I have violated this type of validity in this example.
Statistical validity?
I want to figure out how students at OSU feel about generative AI, but I can't sample every single student, because I don't have the time or money. OSU gives me a list of the student ID numbers of active students, and I have Excel randomly pick 1000 of those numbers, which is an example of this kind of sampling.
Simple random sampling
This kind of explanation seeks to explore all the reasons one specific event occurred.
Idiographic explanation
I am a scientist who is sure (or at least pretty sure) that there is an objective reality, and that it is knowable (or at least, mostly knowable).
Positivist or post-positivist
I am exploring participation in and attitudes toward ziplining amongst tourists in Costa Rica. I give surveys to participants at specific ziplining companies as well as tourists in popular tourists areas. This is my unit of observation.
The individual visitor
I want to explore people's attitudes toward sharks, so I asked the question, "How much time do you spend at the ocean?" In this example, I have violated this form of validity.
Measurement validity
I am doing my Master's thesis research in Grand Teton National Park to explore user experiences at Leigh Lake. Every day, I pick three numbers between 0 and 59, which are the minutes on the hour that I will intercept the next guest, which is an example of this kind of sampling plan.
Systematic sampling
This kind of reasoning starts with a pre-existing theory and uses it to explore similar relationships in a new context.
Deductive reasoning
I am a scientist who believes that reality is a construct, and that these constructs favor those who are already in positions of power. As such, my research seeks to empower communities with less agency, which is why this is my research paradigm.
Critical theory
This is the level of measurement of this variable: Your respondent's attitude toward e-bikes, on a Likert scale from "Strongly approve" to "Strongly disapprove."
Ordinal variable
What is a confounding variable?
A variable that explains the relationship between two variables that was not accounted for in the survey design.
I am hoping to survey a large community that has just been affected by a wildfire, which is comprised of about 20 neighborhoods. I randomly select eight of the neighborhoods, then randomly select eight of the blocks in each neighborhood to sample, which is an example of this kind of sampling.
Cluster sampling
This kind of explanation seeks to explain a few key reasons why a class of events happens.
Nomothetic explanation
I am a scientist who believes that reality is a construct that changes based on its context. This is my research paradigm.
Constructivism/interpretivism
This is the level of measurement of this variable: the number of bears your respondent has seen on their property in the last ten years.
Ratio variable
I want to explore Oregonian's rates of participation in activities like birding and wildlife watching, but I am very busy teaching FES422, so I only visit parks in Corvallis. I have violated this kind of validity by doing this.
External validity
I am interested in Oregonians' attitudes toward hunting barred owls, but I especially want to explore the differences in attitudes between folks in urban areas and folks in rural areas. I know 80% of Oregonians live in urban areas, so I gather a random sample of 800 Oregonians in Multnomah County (Portland) and 200 Oregonians from Umatilla County (Pendleton), which is an example of this kind of sampling strategy.
Stratified sampling
This kind of reasoning starts from observations and thorough exploration of a concept to create a theory to explain a particular phenomenon.
Inductive reasoning
I am a scientist who will take your constructivist paradigm and say it right back to you. You think reality isn't knowable--but what is knowing? You think power dynamics should be explored--but whose power? I am a lot of fun at parties because this is my research paradigm.
Post-modernism
I am exploring participation in and attitudes toward ziplining amongst tourists in Costa Rica. I give surveys to participants at specific ziplining companies as well as tourists in popular tourists areas. This is my unit of analysis.
The individual visitors
I am conducting an experiment to test whether an environmental education program improves children's connectedness to nature. Since my children are enrolled in this program, and I talk to them about my hopes for their time in this program in the evenings, I have violated this form of validity.
Internal validity
This is the group of people from whom I select my sample, which should be as close to the total population I'm representing as possible.
Sampling frame.
This is the difference between probability sampling and non-probability sampling.
Probability sampling uses probability theory, or the idea that randomly selecting participants will result in the most representative sample, while non-probability sampling does not randomly choose participants, but samples them purposively or conveniently.
I am a scientist who is literally so tired of the fighting amongst folks who have different paradigms. I don't care what paradigm makes sense to you; I care what makes sense for exploring each problem. This is my paradigm.
Pragmatist.