What is Science
Key features of “Science” (DEFINE)
Willingness to change with new evidence, Ruthless peer review, takes account of all new discoveries, invites criticism, verifiable, results, limit claims of usefulness, accurate measurement
Experimental studies (CONNECT/COMPARE)
This is when a researcher introduces a procedure and then a result is observed. An example of an experiment could be if a researcher wants to test the effects of coffee on people who have gotten less than 5 hours of sleep over a month-long period.
Example: An experimental study could be researchers conducting a study to determine if the amount of caffeine someone consumes makes them focus better. The independent variable is the amount of caffeine someone has and the dependent variable is how well they focus. Participants split up into 4 groups in which each group is to drink a certain amount of caffeine from coffee 30 minutes after they wake up. The amounts range from 0 mg to 600 mg. They then must do a thought-intensive exercise for 2 hours. The participants do this every day for a week. The researchers recorded their findings of how the amount of caffeine consumed (the independent variable) affected the participants' focus during the thought-intensive exercise.
Independent variable (DEFINE)
The independent variable is manipulated by the experimenter and it has at least two different conditions. An independent variable will have at least 2 conditions, a “treatment” condition and a “control” condition.
Correlation (DEFINE)
This describes the association between two variables. When one variable increases or decreases, the other variable also increases or decreases. It is important to note that this does not equal causation.
Correlation coefficient is represented by (r), range from -1 to 1 with a value close to -1 representing a strong negative correlation and a value close to 1 representing a strong positive correlation. A value of 0 or close to zero indicates no correlation or linear relationship.
Sample (DEFINE)
A sample is a subset of a population of interest selected for a study with the aim of making inferences. While random sampling is when each sample has an equal probability of being chosen, stratified sampling is when samples are divided into subgroups called strata based on shared characteristics. What makes a good sample is when the subset is representative of the intended population in which the study is aimed.
Key features of Pseudoscience (CONNECT/COMPARE)
Fixed ideas, Peer review, Selects only favorable discoveries, Sees criticism as conspiracy, Non-repeatable results, Claims of widespread usefulness, “Ball-park” measures
Example: There is a new pill being sold that is allegedly a “weight loss pill”. This pill claims to be backed by science and is advertised well to its intended population, but it has no scientific evidence that the pill promotes weight loss.
Quasi-experimental design studies (CONNECT/COMPARE)
When groups of individuals are assigned to random treatments, not random assignment
Example: Researchers want to see how the amount of hours of exercise an individual gets each week affects their energy levels. In this study, the researchers couldn’t just divide the groups with a randomized control trial, because the individuals that they are studying have different backgrounds with exercise. Some are sedentary most of the day and some work out every day. So the researchers would have to divide the groups so that individuals with similar exercise habits are together, and then everyone in each group would be randomly assigned an exercise regime for the study. This way it is not a true randomized control study. The researchers attempt to control any confounding variables by using the quasi-experimental design.
Dependent variable (CONNECT/COMPARE)
The dependent variable is measured by the experimenter and it is used to determine the effect of the independent variable. Multiple dependent variables can be measured in order to see how they are affected by the independent variable.
For example, in Albert Bandura’s Bobo Doll Experiment, they are studying whether children will copy the behavior of an adult that plays with a bobo doll. One group is in a room exposed to an aggressive model of behavior with the doll and the other group is in a room exposed to a non-aggressive model of behavior with the doll. In this example the dependent variable is the children's behavior when exposed to a model of behavior. In order to measure their behavior, the child stays in the room for 20 minutes, and their behavior is observed through a one-way mirror. During 5-second intervals, their behavior was observed and rated.
Causation (DEFINE)
This is when one event occurs because of the effect of another event and shows the relationship between cause and effect.
3 requirements for this:
Covariation: We must observe a relationship between the independent and dependent variables
Time-order relationship: The presumed cause precedes the effect
Elimination of plausible alternative causes: Using control techniques, we rule out other possible causes for the outcome.
Simple random sample (DEFINE)
This sample type allows each member of a population an equal chance of being chosen because it is completely random.
Surveys (CONNECT/COMPARE)
Surveys are used to describe opinions, attitudes, and preferences using a predetermined set of questions. They make predictions about behavior using correlations. They are more cost-effective, quick, and easy. However there can be a misunderstanding of questions, you get limited insight, people could answer dishonestly, it could not be a representative sample, and they only show correlation and not causation.
Example: A teacher asking their students to fill out an anonymous form asking questions about how they thought the semester went. This lets the teacher get a quick insight of how their teaching methods are working but many students may not be completely honest and some students may not even complete the survey.
Confounding variable (also known as the “third variable problem”) (CONNECT/COMPARE)
Confounding variables undermine your ability to draw causal inferences. An example of a confounding variable is a placebo effect.
Example: One confounding variable can be the placebo effect. Let’s say that the subjects in an experiment for a new pill to improve focus know if they get the actual pill or not. This could affect the results of the study because the people who received the pill subconsciously might think it helps, resulting in better focus. On the flip side, it could make the people who did not receive the pill subconsciously think that their focus is worse than the other groups, making them focus less.
Population (DEFINE)
This is the group that you want to draw conclusions from. It is the group of people that the study is being aimed at. We want the sample to represent the population.
Stratified random sample (CONNECT/COMPARE)
This is when every member in the population doesn’t have an equal chance of being chosen and instead the population is divided into smaller subgroups based on shared characteristics and then participants are randomly selected from within their groups.
Example: we want to know the university undergraduate students’ ratings of their experience living on campus in comparison to those commuting. For our sample, we will randomly select 25% of our sample from Freshmen, 25% from Sophomores, 25% from Juniors, and 25% from Seniors, in the hopes that our sample will be a good representation of the general population of undergraduate students at the University of Minnesota.