CMS 348 Notes

o Why Study Methods?

1. To do research

2. To evaluate research 

o What is Science?

• The pursuit of understanding and gaining knowledge.

• An inquiry using a systematic method.

• Helps generate knowledge.

• Science is Latin for knowledge.

• Social science is about humanity and the way people think

• One of the main objectives is to learn about things that they want to study or find interesting.

• Science doesn't tell us about morality or what is good or bad. Science tries to be impartial. 

• Science's purpose is not to settle moral debates.

• Social science uses statistics

• Social science doesn’t prove things to be true, they accept things to be true after running a series of tests. 

o Probabilistic Patterns/Social regularities:

• Finds general patterns

• Tautology 

01/23/2025

o 3 Methods: 

1. Experimentation:

• Can prove causality

• Has control and treatment groups 

• Artificial settings (fake/scripted)

• Random Assignment, you randomly assign the conditions

2. Survey Research:

• Correlational answers, not manipulating anything

• Asks for self-reported behavior 

3. Content Analysis:

• Actual past behavior 

• Correlational answers 

• No human participants

- There is no best method, it just depends on what you are trying to accomplish/display. 

- It is more powerful when you use multiple methods.

o What is theory: 

• An account that explains how and why things occur. 

o Goals: Explain, Describe, Predict, Control 

• Primary Goal is to explain

• Hypothesis: a specific prediction about the relationship between at least two variables

• Social science does not have set methods.

o 2 Forms: 

• Null Hypotheses: Zero, nothing.

• Alternative Hypotheses: 

  • You want to reject the null hypothesis. 

o Nomothetic v. Idiographic Theories:

• A way in which we construct our theories in which we construct our theories comprehensively or less comprehensively

• Nomothetic: broad encompassing explanations that account for a lot of people and a lot of different situations

• Idiographic: explaining specific situations. Concrete and limited in focus, less quantitative, much more limited.

o Deductive v. Inductive 

- Where are you starting in the research process and are you prioritizing observations over theory 

• Deductive: prioritizing theory leading to observations 

• Inductive: prioritizing observations leading to theory 

- Reciprocal in nature

01/28/2024

General Pattern of research 

                            ↱Theory⤵

General Conclusions       Hypothesis

                       Observations ⤶ 

• Will you go right to left or up and down

o Ethics: 

• There was a period of time when they were no ethical guidelines for research and no oversight.

• Government said that there needed to be some ethical procedures put into place. 

• The main ethical goal in research is doing no harm. 

o Research Ethics:

o Do No Harm  

• Requires a Cost-Benefits Analysis (must weigh the harm with the benefit that society could gain) (If I do produce harm, that harm will be offset by the good the harm does)

o  Deception

• Yes, you can use deception in research

• Allows us to make artificial information seem legitimate.

• Debriefing: If you use deception, the researcher has to disclose at the end that they were lying. Usually immediately after the subject, people have the option to withdraw their participation after the debriefing. 

o Privacy:

• Anonymity: not collecting identifying information. 

• Confidentiality: Collects identifying information, but we are not going to tell anyone your identity.

• Easiest way to maintain privacy is to have the study be anonymous. 

• Maintain identification by immediately separating the identifying info from the identity of the participants of the study. 

o The Internal/Institutional Review Board (IRB):

• They have specific knowledge in research and what is required for ethical research

• Very hands on and leads to a lot of feedback

• All research must go through the IRB

• As soon as you start studying people you have to get approval from IRB

02/04/2025

o Conceptualization vs. Operationalization:

• Conceptualization: Theoretical, Abstract, RQs

• Operationalization: Measurement, Concrete, Results

o Variable: any characteristic or factor that can take on different values and is measured or manipulated in a study

• Helps determine the kinds of concepts that we are measuring 

• The variable will correspond to concepts. 

o Levels of Measurement: 

• As you move downward you gain additional information.

o Reliability and Validity: 

• Reliability: Consistency, when you ask someone a measure you are going to be able to get the same answer.

• Ways of testing: test-rest, internal consistency,

• Validity: Accuracy, less statistical, how accurate is my measure 

• Does your conceptualization and measure match

• Ways to test: 

• Face validity: does my measure make sense, is it logical.

• Content validity: capturing the breadth. If you say your concept is the dimension but your measure only captures 10% of the dimension. 

• Both about the fit between your concept and measure

• Predict validity: if your 

SAT score does not predict your completion of highschool.

02/06/2025

o Units of analysis: entity about which you want to generalize.

• Most of the time the unit of analysis will be people.

• Can be groups of people.

o Correlation: association between at least 2 variables.

• Correlation coefficient: Ranges from -1.0 to 1.0 |0.-1|

• Positive direct 

• Negative inverse

o Magnitude: how close the scatter is to the fitline

o Direction: Is the slope/fitline going downward or upward.

o Causality:

• First you need a correlation

• If there is no relationship between the two variables you do not have causality.

• Second you need Time order: the cause must precede the effect

• Non-spuriousness: cannot be explained in terms of some third variable. The results cannot be spurious 

• Cross-sectional vs. longitudinal: are you collection data at a single point in time, one date collection effort (cross sectional) or multiple points in time (longitudinal)’

• Cross-sectional sample: Fill out questions about how much they are on social media and how it makes them feel and then you’re done.

o Types of longitudinal designs:

• Trend: it tracks the same concept over time and at each data collection point is tracking a new sample. Ex: Same thing as cross-sectional, but you do it over and over again with different samples, a different set of people. 

• Cohort: examines specific subpopulations, or cohorts, as they change over time. Ex: age group, millennials. Ex: Just interested in Gen Z, it's the same as the trend study but it focuses on a specific age group and your sample will be from the same generation.

• Panel: same set of people time and time again.  Ex: Get a sample of people and make them do the study 3 times in 3 years, measure individuality, and change over time. 

• Longitudinal survey research analysis can be a means for causality.

• More sociological: trend and cohort

• Development, psychological: panel. 

• If its a measure its a operationalization, if its an opererzation it is not necessarily a measure.