SOCYA348 Exam 1 Study Guide

What are the different types of research? Give examples of them.

• Exploratory: Explores a groundbreaking or unstudied topic. More inductive and often uses smaller samples.

  • EX. Dr. Mason’s study on duck dynasty’s effect on attitudes of duck hunters

• Descriptive: Designed to describe how a topic looks. Not trying to find cause or relationship. Summary statistics

  • EX. Public health survey on heart disease

• Explanatory: Designed to explain “why” through theory based hypotheses. More deductive.

  • EX. Study on the relationship between religiosity and abortion attitudes

What are the three criteria for causality? Explain them.

1. Temporal/Time Order: the cause must occur chronologically before the expected effect. Better established with longitudinal studies.

2. Association/Correlation: variables must show a significant relationship with each other, positive or negative.

3. No Spuriousness: There cannot be a hidden 3rd variable that explains the relationship between the 2 variables

Know how to operationalize a variable. How do we come up with our measurements?

• Operationalizing a variable means taking a concept (abstract elements) and specifying what indicators will be used for your variables.

• We come up with our measurement by figuring out the question plus options

  • EX. We want to operationalize age.

    • Question: What is your age?

    • Options: young adult, middle age, old, etc.

Know the difference between inductive and deductive methods. Give examples.

• Inductive: Starts with observations, the identifies a pattern, derives tentative hypotheses, and proposes a grounded theory

  • More stages, all theories began with inductive research

• Deductive: Starts with general knowledge through a theory, derives hypotheses from said theory, collects evidence and observations to explore hypotheses, and supports or rejects the hypotheses.

  • Traditional image of science

What is grounded theory?

A theory derived from data during a study setting and based on collection and analysis of real world data.

Identify and give an example of several different units of analysis. Illustrate the ecological fallacy using your examples.

• Individuals: Which students are most likely to be addicted to their cell phones?

• Groups: Do certain types of social clubs have more cell phone addicted members than other clubs?

• Organizations: How do different colleges address the problem of addiction to cell phones?

• Ecological fallacy: making inferences about individuals from information about groups that might not be exclusively compromised of those individuals

  • EX. Social clubs with higher cell phone addiction also have more athletes=> Athletes are more addicted to their cell phones

Know the characteristics of cross-sectional studies and longitudinal studies. What are the advantages and disadvantages?

• Cross-sectional: Data is collected at 1 point in time, aka snapshot.

  • Adv: Lower cost and time investment, large samples, low attrition, can be anonymous

  • Disv: Harder to infer causation because no time order information

• Longitudinal: Data is collected at 2 or more points in time; trend, panel, and cohort studies

  • Adv: Tracks changes across time, temporal order can be established, allows for greater analysis

  • Disv: High cost and time investment, not anonymous if panel, panel mortality & conditioning

List and discuss the various forms of longitudinal research. In addition, give a potential scenario where one may or may not be used. What are the benefits and shortcomings?

• Trend: data gathered 2+ times with a NEW sample

  • Study looking at SAT scores of 11th graders in SC

  • Adv: Less expensive than panel, can be anonymous, no attrition,

  • Dis: Can’t compare individuals across time, time order not as clear as panel

• Panel: data gathered 2+ times with the SAME sample

  • National Crime Victimization Survey

  • Adv: Best for causality, can compare individuals across time

  • Dis: High cost, panel mortality and conditioning

• Cohort: studies a cohort, could be trend or panel

  • Study looking at the experiences of the USCA graduating class of 2026

    Adv: Tease out period effects and life course events

  • Dis: Depends on if trend or panel

What are descriptive statistics?

Numerical summaries that describe the main features of a dataset. Measures of central tendency (mean, median, mode), variability (range, variance), and distribution (frequency).

Know how to calculate the range, mean, median, and mode. When is it appropriate to use them? Which are best for things that cannot be ranked/ordered? What about when the data are highly skewed or large variances?

• Range: The largest/highest number in the set minus the smallest/lowest number

  • Best for data with no outliers

• Mean: Add all values and divide by the number of values

  • Best when data is highly skewed or large variances

  • Best for not skewed data

• Median: Value that sits above 50% of the set and below 50% of the set

  • Best for skewed data

• Mode: Value that appears most often in the set

  • Best for nominal data

What are inferential statistics? What are they inferring?

• Body of statistical computations relevant to making inferences from findings based on sample observations to some larger populations.

• Inferring information about the population

Know how to recode a variable in SPSS.

Know how to run frequency and contingency (crosstab) tables.

What does statistically significant mean? Specifically, what does it mean to say we have a result where p<.05? What does the percentage mean?

• Statistically significant means that the observed relationship between variables is unlikely to have occurred by random chance.

• When we have p<0.05, we are 95% confident that the relationship in our sample is generalizable to the population and did not occur by chance.

• If the p-value= 0.02, this means there is a 2% chance that our relationship occurred by chance.

How does the Chi-square relate to expected and observed counts among the variables? In other words, is the test assuming there is a relationship, or there is not a relationship between the variables?

The Chi-square assumes there is no relationship between the variables. The Chi-square expected counts calculates the counts if the null hypothesis is correct. The observed counts are the counts actually seen in the data.