Data Analysis for Decision Making

Running header: Data Analysis for decision making 1
Data Analysis for Decision Makers
Name:
Institution:

Data Analysis for decision making 2
Introduction
There is no doubt the success of any organization depends on the ability to make
informed decisions. As a result, industries, companies, organizations, and facilities have adopted
various methods that aid in decision-making, which include data analysis and interpretation.
Therefore, the following report seeks to exhibit how various measures, tests, and analysis can be
used to analyze data and make informed decisions. The report uses the data collected in the
tourism sector the exhibits the hospitality and accommodation services provided by various
restaurants.
The dataset incorporates 50 observations (entries) and ten variables, which include Bord
Failte ID, category, star rating, number of beds, number of guests, numbers of guests taking
breakfast, number of guests taking dinner, length of stay for current guests, revenue, and average
time spent by a guest. Notably, Bord Failte ID uniquely identifies the observation and category
exhibits the type of service offered, which include bed and breakfast (BB) and hotel services. On
the other hand, star rating shows the level of service provided by respective categories in the
hotel industry whereby the ratings are from 2 to 5; moreover, the length of stay exhibits the
number of days the current guest will spend in the hotel or BB.
Random Variables
Notably, there are two categories of random variables, which include continuous and
discrete random variables. Discrete are variables that assume finite values, whereas continuous
are variables that assume infinite values within a specified limit. As evident, the dataset contains
both discrete and continuous variables, whereby the discrete comprise of star rating, number of
beds, number of guests, numbers of guests taking breakfast, number of guests taking dinner,

Data Analysis for decision making 3
length of stay for current guests. On the other hand, continuous variable comprises of revenue
and average time spent by a guest.
Probability Distributions
A probability distribution is a function that exhibits the possible outcomes that a random
variable can assume in a random process with the associated probability of occurrence. Notably,
one of the critical components of a probability distribution is the random variable, whereby
discrete random variables assume a probability distribution known as the Probability Mass
Function (PMF) whereas continuous assumes the Probability Density Function (PDF). There are
various forms of probability distributions, such as uniform, Bernoulli, Binomial, and Poisson
(PMFs), and normal and exponential distribution. For instance, length of stay by the guest is
unpredictable thus possess low probabilities; as a result, the efficient probability distribution to
represent this variable is the Poisson distribution. Therefore, Poisson can be used to determine
the probability that a given customer will stay in the hotel for specific days.
Notably, among the PDFs, the normal distribution is the most common whereby it
represents the behavior of most occurrence or events. There are various characteristics associated
with the normal distribution, such as the equal mode, mean, and median, the curve of the
distribution is doom-shaped and symmetrical about the mean, and total area under the curve is 1.
Descriptive statistics
The primal analysis for any data set is the descriptive statistics that aid in the exhibiting
the fundamental characteristics of the data. Notably, there are two categories of descriptive
statistics, which include the measures of central tendency and dispersion.
Measures of Central Tendency

Data Analysis for decision making 4
A measure of central tendency is a descriptive summary of a given dataset through a
unique value that reflects the center of the data distribution (ABS, 2013). Notably, the three
measures of Central location, which include mean, mode, and median. Among the three, mean
(arithmetic mean) is the most commonly used measure whereby it represents the sum of
observations divided by the total number of the observations or entries (Manikandan, 2011).
Besides, there are other types of mean, such as the geometric and harmonic means. The harmonic
mean is determined by getting the reciprocal of the arithmetic mean of the observations, whereas
the geometric mean is the arithmetic mean of all the observations taken on a logarithm scale
(Manikandan, 2011).
Notably, there are various advantages associated with mean, which include incorporation
of all the observation, can be used for both continuous and discrete numeric data; however, there
are numerous disadvantages linked to mean, such as sensitivity to extreme values (outliers)
(Manikandan, 2011). Moreover, there are multiple standard statistical notations used in the
computation of mean whereby the population mean is denoted using (μ) mu, whereas the sample
mean is denoted by x-bar (x̅). As a result, the calculation of arithmetic mean is given by;
μ=(∑X/N). For instance, the arithmetic mean of revenue generated in one night among the 50
hotels is computed by adding all the revenue generated divided by the number of hotels (50).
On the other hand, the mode exhibits the most frequently occurring observation (value)
within a dataset whereas the median defines the middle value within a dataset that may be
arranged in either ascending or descending order (ABS, 2013). Notably, in cases whereby the
total number of observations in the data is an even number, then the median is calculated by
getting the average of the two middle values. As evident, the mode of the length of stay among
the current guests is 2, whereas the median is 2 days. Notably, the median is not affected by