SOCI 3311 (Sociology Statistics) Assessment 1

What is a Measure of Central Tendency?

1. Numbers that describe what is "average" or typical of the distribution. 2. They indicate where cases tend to cluster in a distribution or what is typical or common in a distribution.

What type of variables can Median be used with?

1. Ordinal. 2. Interval-ratio.

What "Central Tendency" is most commonly used?

Mean.

Frequency

The number of times an event or values occurs.

How do you find the Medain in an Odd Size Sample?

(N+1) / 2. This will find the LOCATION of the median.

If in a certain sample, about 42 percent of respondents reported that they did not attend sports events regularly. What proportion of respondents did not attend sports events regularly?

0.42

In a sample of 350 people, 100 are white, 50 are black and the rest are Hispanic. What proportion of the sample is Hispanic?

0.571

What type of variables can Mode be used with?

1. Nominal. 2. Ordinal. 3. Interval-ratio.

Value Labels

Descriptive labels for the attributes of a variable.

Variables

1. A property of people or objects that take on 2 or more values. 2. Must include categories that are both Exhaustive and Mutually Exclusive. 3. EX: Social class, age gender, income...

Attributes

1. Answers to the variables. 2. EX: 65 years old, male, blond, healthy...

Interval-ratio

1. Based on real or relative numeric values. 2. Age and Income.

Ordinal

1. Can be ranked from high-to-low but NO numeric values. 2. Social Class.

Nominal

1. Cannot be ranked. 2. Political party, race, religion.

Find the Dependent and Independent variable: "Educational attainment may cause a persons salary to increase."

1. Dependent = Salary. 2. Independent = Educational attainment.

What type of variables can Mean be used with?

1. Interval-ratio

Units of Analysis

1. Level of life we are studying. 2. The levels of social life on which social scientists focus. 3. EX: Family, Organization, City.

What are the 3 ways to measure "Central Tendency"?

1. Mode = most frequent. 2. Median = middle case. (N+1) / 2. 3. Mean. Average. (Sum of x / N)

What are the Levels of Measure?

1. Nominal. 2. Ordinal. 3. Interval-ratio.

Independent Variable

1. The Cause 2. IV usually occurs earlier in time???

Dependent Variable

1. The Effect. 2. DV is always the property you are trying to explain. 3. The Object of the research.

Operationalization

1. The process of devising a specific measure for a variable that we have conceptualized. 2. EX: Biological sex, age, race, religion, marital satisfaction...

What are the Properties of Medians?

1. They always exist (always can find). 2. They are always unique. 3. The are always at the center. 4. They are not easily affected by extreme values (robust).

What are the 5 columns of Frequency Tables?

1. Value Labels. 2. Frequency. 3. Percent. 4. Valid Percent. 5. Cumulative Percent.

In a sample of 660 respondents, men account for 45% of all observations in the sample. What is the total number of women in the sample?

363

In a sample of 350 people, 100 are white, 50 are black and the rest are Hispanic. What percentage of the sample is Hispanic?

57.1%

Histogram

A bar graph of an Interval-ratio variable in which the height of each bar reflects the frequency of each variable.

Percentage (%)

A relative frequency obtained by dividing the frequency in each category by the multiplying by 100. P = (f / N) x 100

Proportion (P)

A relative frequency obtained by dividing the frequency in each category by the total number of cases. P = f / N

Frequency Table

A table reporting the number of observations falling into each category of the variable. Example: (Variable = Frequency). 1. Married = 40,565,210. 2. Never Married = 61,256,200. 3. Separated = 2,563,100.

What does "average" tell you?

Allows us to summarize, describe, and predict events in our daily lives.

What is a characteristic of Normal Distribution?

Bell shaped curve.

What are Outliers?

Cases with values either higher or lower relative to the typical patterns in a distribution. They tend to skew the data. Effect the Mean but not the median or mode.

Empirical Research

Easily measured objectively. "Are women paid less than men for the same type of work?"

Ungrouped Data

Every value in the dataset is listed, usually in a row or column. Usually only Ungrouped if it is a small dataset. Example: 2 5 8 3 2 1 5

How can you easily find the Median?

Find the Cumulative Percent at 50%

How can you easily find the Mode?

Find the category or score with the largest Frequency of Valid Percentage.

Identify the independent and the dependent variable, as well as the unit of analysis, in the following hypothesis: Documented immigrants in the United States are less likely than undocumented immigrants to use and depend on social services (e.g., welfare, food stamps, etc).

Independent = Immigrant Status. Dependent = Dependence on Social Services. Unit of Analysis = Individual humans.

Identify the independent and the dependent variable, as well as the unit of analysis, in the following hypothesis: On average, the annual household income of female headed households is about 25 percent less than the annual household income of married couple households.

Independent = Marriage Status. Dependent = Household Income. Unit of Analysis = Household.

What is the median, mode, sample size, and mean for the following scores: 8, 10, 2, 8, 12, 13, 7?

Median = 8. Mode = 8. Sample Size = 7 Mean = 8.57.

How do you find the Medain in an Even Size Sample?

N / 2. This will give you 2 different values. If the data is ORDINAL, report both values/categories. If the data is INTERVAL-RATIO, report the average of the 2.

Why do researches use percents?

Percent is translated from the raw frequency to better understand the meaning of the frequency. This helps us compare the variables.

What do you report if the data is Skewed?

Report the Median.

Cumulative Percent

Shows the percent of cases above or below a certain point or attribute. Comes in HANDY with Ordinal or Interval-ratio variables.

What is Skewing?

Significantly pulling the distribution in one direction.

Valid Percent

Standardizes frequencies by expressing them as the number of times an attribute occurs per 100 cases. Based only on those cases that provided data (DO NOT INCLUDE MISSING DATA).

What is an example of Non Empirical Research?

Subjective. "Is racial equality good for society?"

Is the Mean affected by unusual values (outliers)?

The mean is seriously affected by outliers.

Can Outliers be removed from the data for analyses?

Yes, but it must be justified.

If the data is symmetric, would the mean, median, and mode be similar?

Yes.