Business Analytics Exam 3

percentile

a measure that tells you what percentage of the total frequency scored at or below a given value

percentile rank

a percentage of scores that fall at or below a given score

quartiles

values that divide a ranked dataset into four equal parts

First Quartile (Q1)

the median of the lower half of the distribution

Second Quartile (Q2)

the middle value of the entire distribution

Third Quartile (Q3)

the median of the upper half of the distribution

Which quartile is the 25th percentile in?

First Quartile

Which quartile is the 50th percentile in?

Second Quartile

Which quartile is the 75th percentile in?

Third Quartile

interquartile range (IQR)

the spread of the middle 50% of the data

formula for IQR

Q3-Q1

range

the difference between the max and min

formula for range

max-min

five number summary

the core components required to build a boxplot

what are the components to the five number summary?

min, max, Q1, Q2, Q3

outliers

measurements that are very different from the rest of the data

whiskers

dashed lines that plot the range of the data

right skewed

tail stretches to the right and the median is closer to Q1

symmetric skewness

balanced evenly, median is centered between Q1 and Q2

left skewed

tail stretches to the left, median is closer to Q3

68-99-99.7% rule

68% of the data falls within 1 standard deviation of the mean, 95% within 2 standard deviations, and 99.7% within 3 standard deviations

analytics

a process that examines data using a variety of statistical tools to discover, interpret, communicate, and apply meaningful pattens toward effective decision making in the business workplace

data literacy

the ability to read, understand, and interpret data types and data collection methods, and then accurately communicate data to drive informed business decisions

descriptive analytics

using historical data to summarize past and current events, directly answering the question “what happened?”

diagnostic analytics

goes beyond descriptive analytics by digging into root causes, patterns, and relationships between variables. determining “why did this happen?”

predictive analytics

examines data to discover and predict what might happen in the future

prescriptive analytics

studies data to determine and recommend the next best actions

business problem

identifying what is being investigated because a business needs to make money or address a specific problem

business problem guidelines

must address the objective, focus on one topic, be measurable, have obtainable data

data collection

process of gathering, measuring, and recording information to analyze performance, answer specific business questions, and guide strategic decision making.

actional business decisions

using presentation tools (tableau dashboards, reports, graphs) to tell an accurate story and drive business actions

quantitative

numerical data that can be measured

continuous data

quantitative data that can take on any value within a range

example of continuous data

weight, height, time

discrete data

quantitative data that consists of separate, countable whole numbers

example of discrete data

number of children, number of items purchased

qualitative data

data that is non numerical. used to describe things and categorize based on traits or characteristics that can be counted.

nominal data

qualitative data where there is no meaningful ordering of categories

example of nominal data

gender, car color, product names

ordinal data

qualitative data where order does matter

example of ordinal data

class rank: freshman, sophmore, junior senior

is a product name/brand qualitative or quantitative?

qualitative

is a customer review rating of 4 stars quantitative or qualitative?

quantitative

is price ($19.97) quantitative or qualitative?

quantitative

point of sale

accepts customer payments and tracks sales and inventory for online or physical stores

clickstream data

tracks a users navigation behavior and the pathway/clicks they take through an online website or search engine

social media data

captures a users interactions, opinions, thoughts, and virtual network sharing.

sensor data

uses a device that detects environmental changes and converts them into a measurable digital signal.

confidential

identification is coded but the real identity is known by someone in charge

anonymous

no identification is known by anyone

population

the endite group of individuals or items you want information about

sample

a smaller subset of the population actually collected to analyze and draw conclusions/inferences about the whole population

random sampling rule

choosing a sample so that everyone in the population has an equal chance of being chosen

biased samples

if you don’t use random sampling your data favors certain outcomes

convenience sampling

selecting people/things just because they are easy or readily available

voluntary response sampling

self selected sample where people choose to participate (usually because they have a strong opinion)

what causes undercoverage and why?

biased sampling methods because everyone in the population was ever on the list to be considered.

what sampling methods are biased?

voluntary and convenience

simple random sampling (SRS)

every individual has an equal chance, and every group of size n has an equal chance of being selected

stratified random sampling

divide the population into non overlapping groups called strata (based on a specific characteristic like gender, class rank, age group), then take a random sample from every single group

what error occurs if your strata doesn’t relate directly to your business problem?

measurement error

systematic sampling

make a list of the population, pick a random starting point, and then sample every nth element (every 10th car, every 5th product off a manufacturing line)

multi stage cluster sampling

divide the population into clusters (groups), randomly select whole clusters and sample everyone inside those chosen clusters.

what sampling method is often followed by a stratified step to rescue lost characteristics?

multi stage cluster

what sampling methods are unbiased?

simple random, stratified, systematic, multi stage cluster