Data Analysis for Business Exam

What are the four steps of the SOAR analytics model?

S - Specify the question, what business problem are we trying to solve?

O - Obtain the data, gather relevant reliable information.

A - Analyze the data, look for patterns and calculate statistics.

R - Report the results, communicate findings so people can make decisions.

What does DIKW stand for and how is it used by businesses to improve efficiency, effectiveness, processes and gain a competitive advantage?

D - Data, raw facts.

I - Information, data organized to have meaning.

K - Knowledge, understanding what the information tells us.

W - Wisdom and using that understanding to make a decision.

What are the AI basics in data?

Artificial intelligence is the broader field.

Machine learnings learns patterns from data.

Deep learning uses layered neural networks.

Gen AI creates content and can produce incorrect information so you have to verify it’s output.

What does SCOPE stand for in terms of making key decisions involved in obtaining and preparing data?

S - Sources, internal data that comes from within a company or external data from outside the company like census, social media, stocks or market reports.

C - Collect, decide what data is needed and how to obtain.

O - Organize, understand how data is arranged such as tables and databases.

P - Prepare, check, clean and handle missing values before analysis.

E - Evaluate and ethics, check personal information and use data responsibly.

What are the 5 important types of data and the four Vs of data?

S - Structured data, organized in tables.

U - Unstructured data, for example pictures, blogs and tweets.

S - Semi structured data, has some organization but not a rigid table format.

T - Text data includes tweets and documents.

R - Raw data gives the analyst flexibility to perform their own analysis such as totals, counts or averages.

Four Vs of Big Data:

Volume, amount of data.

Velocity, speed at which data is generated or processed.

Variety, different data formats and types.

Veracity, data trustworthiness and quality.

What are the four data types?

N - Nominal, categories with no natural ranking.

O - Ordinal, categories that can be ranked.

I - Interval, numerical data with equal intervals.

R - Ratio, numerical data with equal intervals and can be negative.

Discrete data is countable whole number data.

Continuous data is values within a range.

What are the four sampling methods and key terms of sampling?

Simple random: Everyone has a chance of being selected.

Stratified random: Divide the population into categories, then randomly sample from every category.

Cluster: Example, randomly select two classrooms and survey all students in them.

Convenience: Select whoever is easiest to reach.

Key terms

Population: the entire group of interest.

Sample: a subset of that group.

Census: data collected from the entire population.

Data reduction: Make the dataset more manageable by filtering records. Often cheaper and faster than a census.

What does MARS stand for in terms of identifying the center and spread of a dataset and what are the basic statistics to memorize?

M - Mean, average = sum of values/number of values.

A - Analyze, compare the mean, median and mode.

R - Range, maximum - minimum.

S - Spread, variance and standard deviation

Median is the middle value.

Mode is the most frequent value.

Variance is the spread using squared deviations from the mean.

Standard deviation is the square root of variance.

IQR is Q3 - Q1.

What are the charts and distributions?

Bar chart: compares categories.

Line chart: show changes over time.

Histogram: show the frequency distribution of numerical data.

Box plot: show the median, quartiles, spread and potential outliers.

Scatterplot: show the relationship between two numerical variables.

Pie chart: show parts of a whole.

Distributions

Symmetrical: approximately balanced.

Right-skewed: long tail extends to the right.

Left-skewed: long tail extends to the left.

Normal distribution: symmetrical, bell-shaped distribution.

Uniform distribution: values have approximately equal likelihood across the range.

What are a confidence interval, a hypothesis test, alpha and a p-value?

Point estimate: one sample based value used to estimate a population parameter.

Confidence interval: a range around an estimate; commonly written as point estimate ± margin of error.

Hypothesis: a proposed claim to investigate.

Null hypothesis (\(H_0\)): the baseline claim, often no difference or no relationship.

Alternative hypothesis (\(H_A\)): the competing claim, such as a difference or relationship.

Alpha (\(\alpha\)): the chosen significance level, often 0.05.

p-value: evidence measure used in deciding whether to reject the null hypothesis.

What is the difference between correlation and regression, and how do you interpret them?

Correlation measures the direction and strength of a linear association between two variables.

Correlation coefficient +1: perfect positive linear relationship.

0: no linear relationship.

−1: perfect negative linear relationship.

Regression models the relationship to help explain or predict an outcome, y = mx + b

y: dependent variable.

x: independent variable.

m: slope.

b: intercept.