EXAM 3 Flashcards
7.1 What is Data Exploration?
Define Data Exploration
Also called 1) ex___y data analysis, it is the 2) di____y process of 3) lo___g for something 4) n___ and previously 5) u___n in the data
This is accomplished by looking for 6) p___ns, ou___s, or, more generally, for 7) ins____.
1) exploratory
2) discovery
3) looking
4) new
5) unknown
6) patterns, outliers
7) insights
Define Insights
an 1) obse_____ that might 2) sig____ af___t a business’ 3) de___-m___g.
1) observation
2) significantly affect
3) decision-making.
Remember, decisions are 1) n___ based on 2) d___.
Rather, decisions are informed by the 3) in___s generated from 4) d___
1) not
2) data
3) insights
4) data
This process of generating insights is what 1) dis___ data analysis from the simple act of reporting 2) nu____.
While they might share some tools, it is essential to differentiate between data 3) e___on, int____ion, and rep___:
Exploration: 4) Di__ng in___s.
Interpretation: 5) Con____g and unde____ in___s.
Reporting: 6) Com___ ins___s
1) distinguishes
2) numbers
3) exploration, interpretation, and reporting
4) Discovering insights
5) Contextualizing and understanding insights
6) Communicating insights
The Process of Data Exploration
What’s the 4 step process of Data Exploration?
- Ide___g Q____s
- Id____g D___ Rel___s
- Exp___ D___a Rel____s
- Gen___g In____s
- Identifying Questions
- Identifying Data Relationships
- Exploring Data Relationships
- Generating Insights
The Process of Data Exploration
1. Identifying Questions
Data exploration helps answer accounting 1) que__s such as whether sales and profits are 2) im___g, which products deserve 3) inv____, if 4) b__d d___s are appropriately managed, and more.
An 5) an___ database should provide 6) an___s to both 7) an____d and un____d (un___ed) qu____s.
Once the question is 8) det___, such as the question about whether unit sales are improving in Illustration 7.1, then the underlying data 9) rela___s can be 10) id___
1) questions
2) improving
3) investment
4) bad debts
5) analytical
6) answers
7) anticipated and unanticipated (unplanned) questions
8) determined
9) relationships
10) identified
The Process of Data Exploration
2. Identifying Data Relationships
Define Data Relationships
describes 1) h__ d__ el___s (or v__s) 2) rel___ to each 3) ot___
-But before aspects of data relationships can be analyzed, they have to be 4) id__
1) how data elements (or values)
2) relate
3) other
4) identified
The Process of Data Exploration
2. Identifying Data Relationships
Stephen Few, an expert in data visualizations, differentiates eight foundational data relationships
- No____ co___son
- Di____n
- Dev___
- Ra___g
- Pa__-to-wh___e
- Cor____
- T___e se___s
- Geo___
The data relationship identified in Illustration 7.1 is a time series, which describes 1) ho__ something 2) cha__ o___r ti__ and helps to 3) ide___ pa___s of ch___.
A relationship that has been 4) ide___, whether it is a time series relationship or one of the relationships examined later in this chapter, is ready for 5) ex____n
- Nominal comparison
- Distribution
- Deviation
- Ranking
- Part-to-whole
- Correlation
- Time series
- Geospatial
1) how
2) changes over time
3) identify patterns of change
4) identified
5) exploration
The Process of Data Exploration
3. Explore Data Relationships
While there are different approaches to exploring data relationships, 1) vis____ and s___s are the most common.
Exploration involves 2) sel___ the 3) vis__n or vis___s 4) b___t su__d for 5) ex___g the data 6) rel___s. In Illustration 7.1 a line chart visualizes the time series
1) visualization and statistics
2) selecting
3) visualization or visualizations
4) best suited
5) exploring
6) relationships
The Process of Data Exploration
3. Explore Data Relationships
Keep in mind that 1) t___l-sp___c kno___ is required to create 2) vis___s.
For example, a time series analysis requires knowing how to create line charts.
Business intelligence software such as Excel, Power BI, and Tableau all have powerful tools that visualize data relationships for exploration
1) tool-specific knowledge
2) visualizations
The Process of Data Exploration
4. Generate Insights
The line chart in Illustration 7.1 shows an upward trend of unit sales starting in 2023.
Exploring this insight further would include discovering the 1) s___e of that growth and if the upward trend can be explained by other 2) fa___.
In fact, data exploration is a 3) con___s p___s.
4) Ins___ generate n__w que___, which then generate even 5) m___e ins____. These observations are then 6) int____ and co____ed to stakeholders during the 7) l___t st__ of the data 8) a___s p___ss
1) source
2) factors
3) continuous process.
4) Insights generate new
5) more insights.
6) interpreted and communicated
7) last stage
8) analysis process
Exploring Data with PivotTables
Data exploration 1) inv____s data from 2) di___t an___s to collect 3) in__s.
A widely-used tool for this is the 4) Ex__ Pi___T___e
1) investigates
2) different angles
3) insights
4) Excel PivotTable
The five components used for data exploration with PivotTables are 1) fi___s, v___s, r___s, col___s, and fil___
1) fields, values, rows, columns, and filters
Fields
The Fields area 1) li__ all the data 2) el___s available for 3) exp___n p___es. They can be dragged and dropped to other areas to build data 4) rel____ and fi__r the data.
In Illustration 7.3, the Model and UnitsSold fields are used for exploration.
1) lists
2) elements available
3) exploration purposes
4) relationships and filter
Values
The Values area in Illustration 7.3 (B) represents the 1) n__r or nu__rs to be 2) an__d.
It can be used to 3) e___e data in different ways:
-Drag and drop any field into the Values area and apply mathematical operations such as average, count, or sum to it.
-Create calculated fields.
Examples of accounting-related values that could be analyzed include gross revenue, net revenue, taxes, cost, profit, and more. For HNA, the values in the UnitsSold field are summed, generating the total number of units sold during the 2021–2025 period.
1) number or numbers
2) analyzed
3) exploree
Filters
Using filters further enhances data exploration with Excel PivotTables.
Filters let us determine 1) w___t d__a should be 2) con____ for analysis, and they can be 3) cr___d for any field
1) what data
2) considered
3) created
7.2 How are Data Relationships Visualized for Exploration?
There are two types of data exploration patterns.
Some patterns explore a 1) fou___ data re___p with a 2) s___le vis____n, while others explore data by 3) inte___ da__ re___ps
1) foundational data relationship
2) single visualization
3) integrating data relationships
Data Exploration Pattern 1: Nominal Comparison
Define Nominal Comparison Data Relationship
A data 1) rel___p that 2) com___s the 3) va___s of a 4) cat___l va___le based on a 5) s___d, nu___ variable
1) relationship
2) compares
3) values
4) categorical variable
5) second, numeric
Data Exploration Pattern 1: Nominal Comparison
Define Exploration Structure
A 1) vi___l that 2) d___es the 3) dif___t data 4) e__ts 5) u___d in data 6) exp___ and 7) h__w they are 8) r__d
1) visual
2) describes
3) different
4) elements
5) used
6) exploration
7) how
8) related
Data Exploration Pattern 1: Nominal Comparison
For nominal comparisons, the exploration structure is a 1) n___al variable, which is 2) w___ is being 3) com___, and a 4) nu____ variable, 5) w___ is 6) h__w the values of the 7) n__l variable are 8) co__d
1) nominal
2) what
3) compared
4) numeric
5) which
6) how
7) nominal
8) compared
Data Exploration Pattern 1: Nominal Comparison
Visualizations
Visualizations for nominal comparisons include 1) b___r charts, c___n charts, d___ plots, and loll___p charts.
1) bar charts, column charts, dot plots, and lollipop charts
Data Exploration Pattern 1: Nominal Comparison
Exploration and Insights
A nominal comparison can 1) qu___y ev___e data and collect 2) in___ in__t, which is especially 3) u___l when 4) w____g with a 5) n___ data set.
It allows us to compare the 6) si___s of each category–which is the 7) bi___t, which is the 8) sm___t, if one category is 9) tw__e as big as another, and more
1) quickly evaluate
2) initial insights
3) useful
4) working
5) new
6) sizes
7) biggest
8) smallest
9) twice
Data Exploration Pattern 1: Nominal Comparison
Exploration and Insights
-Any nominal variable, or 1) w___ is being a__d, can be 2) co___d using any numeric variable, or 3) h___w the nominal variable is being 4) a___d.
-Filters can also be applied to both the 5) nom__l variable and the 6) n___c variable.
1) what is being analyzed
2) compared
3) how
4) analyzed
5) nominal
6) numeric
Data Exploration Pattern 2: Distribution
Define Distribution Data Relationship
A data relationship that shows 1) h__w the 2) va__s of a 3) n___c variable are 4) dist___, or sp___ out, by 5) pro___ the 6) dif___ values 7) pr__nt in the data
1) how
2) values
3) numeric
4) distributed, or spread out
5) providing
6) different
7) present
Data Exploration Pattern 2: Distribution
A common scenario, shown in Illustration 7.11, is to create and compare multiple distributions of the same numeric variable based on the different values of a nominal variable.
The dotted line indicates that such a variable is valuable for exploration, but optional when defining a distribution data relationship.
Data Exploration Pattern 2: Distribution
Visualizations
Several visualizations portray distributions, including 1) hist___, v__n plots, and b__-and-w___er charts (or 2) b__lot charts).
Box-and-whisker charts are both 3) po___ and de___d, so they are used here to illustrate a visualization for this pattern
1) histograms, violin plots, and box-and-whisker charts
2) boxplot charts
3) powerful and detailed
Data Exploration Pattern 2: Distribution
Exploration and Insights
Using the exploration structure for a 1) dis___ data relationship, we can 2) q___y examine the distribution of any 3) nu___ field in a data set and use any nominal field in the data set to 4) co__e data distributions.
1) distribution
2) quickly
3) numeric
4) compare
Data Exploration Pattern 2: Distribution
Exploration and Insights
It might be necessary to 1) ex___e the data 2) m___ to 3) be___ u____and the sou__ of the variation.
1) explore
2) more
3) better understand the source
Data Exploration Pattern 2: Distribution
Exploration and Insights
Distribution analysis has many applications, including analyzing employee compensation.
To understand how salaries are distributed, distribution analysis could be used to answer the following questions:
-Are salaries right-skewed? -Are the higher salaries more spread out?
-How do salaries compare across departments?
-How does an organization’s compensation distribution compare to the distribution for other organizations?
Data Exploration Pattern 3: Deviation
Define Deviation Data Relationship
A data relationship that shows 1) h__w a___l values deviate from their 2) refe___ va__s, which are 3) bu___ or fore___ values
1) how actual values deviate 2) reference values
3) budgeted or forecasted
Data Exploration Pattern 3: Deviation
Deviation relationships are everywhere in accounting.
1) Va__e an___s is a good example, such as the difference between actual and standard costs.
The exploration structure for deviation relationships shown in Illustration 7.18 contains the variable being compared and the variables being used for comparison purposes–actual, target, and deviation.
1) Variance analysis
Data Exploration Pattern 3: Deviation
Visualizations
1) Clu___ bar and 2) c___n charts (used in the following example), 3) ga___s, 4) b___t charts, and more can be used for 5) d___ion analysis.
1) Clustered
2) column
3) gauges
4) bullet
5) deviation
Data Exploration Pattern 3: Deviation
Visualizations
Unlike a column chart, a 1) clu___d column chart uses 2) m___e variables for 3) com___n purposes.
While the clustered column chart can make comparisons 4) a___s models, it also provides a comparison 5) w__n a model
1) clustered
2) multiple
3) comparison
4) across
5) within
Data Exploration Pattern 3: Deviation
Exploration and Insights
There are also three nominal variables in this data set to choose from: 1) co___, ty___, or m____l
1) country, type, or model
Data Exploration Pattern 3: Deviation
Exploration and Insights
Different types of 1) in____s can be 2) col___d from 3) dev___ analysis, including when results are 4) h___r or more fa___e than exp____, lo___ or l___s fa___ble than expected, and if targets are 5) unre____.
Once we 6) id___y these insights, the next step might be exploring possible 7) ca___, such as how to reverse 8) unf___e deviations. This is particularly useful when the variances are 9) si___nt.
1) insights
2) collected
3) deviation analysis
4) higher or more favorable than expected, lower or less favorable
5) unreasonable
6) identify
7) causes
8) unfavorable
9) significant
Data Exploration Pattern 3: Deviation
Exploration and Insights
Variances have 1) hi____ been 2) v___ble tools for accountants to indicate issues that should be addressed:
-3) Ineff____ such as 3.1) ine____e in negotiating prices, ill-q____d employees, employees assigned the wrong task, use of l___-quality re___s, or in___e training.
-4) Une____d ch____s such as changes in 5) pr___s, or a 6) dec___e in 7) de___d.
-8) P___r b____ing
1) historically
2) valuable
3) Inefficiencies
3.1) inexperience in negotiating prices, ill-qualified employees, employees assigned the wrong task, use of low-quality resources, or inadequate training.
4) Unexpected changes
5) prices
6) decrease
7) demand
8) Poor budgeting
Data Exploration Pattern 4: Ranking
Define Ranking Data Relationship
A data relationship that 1). or__s the 2) va___s of a variable 3) se___lly based on the values of a 4) se___d variable
Ranks are determined by some 5) qu___, such as 6) hig___, lo___t, f__st, sl___t, as defined by the 7) se___ variable.
The ranking can be displayed in 8) a___ing or des___g order. It is also possible to calculate a 9) r___k exp__
1) orders
2) values
3) sequentially
4) second
5) quality
6) highest, lowest, fastest, slowest
7) second
8) ascending or descending
9) rank explicitly
Data Exploration Pattern 4: Ranking
Visualizations
There are 1) n__ s___ic ra__g visualizations, but many 2) inte___ ranking in____on, including 3) ta___s, b___ charts, and co___n chart
1) no specific ranking
2) integrate ranking information
3) tables, bar charts, and column chart
Data Exploration Pattern 5: Part-to-Whole
Define Part-to-Whole Data Relationship
A data relationship that 1) com___s pa___s to w__es and examines how the 2) dif___ p___s co___e to each other
A part-to-whole visualization should define the 3) nu___er that is the 4) wh__ and 5) h___ the whole should be 6) br___n d__n into 7) pa___.
1) compares parts to wholes
2) different parts compare
3) number
4) whole
5) how
6) broken down
7) parts
Data Exploration Pattern 5: Part-to-Whole
Visualizations
Multiple visualizations can model part-to-whole relationships, including 1) p__ charts, do___ charts, sta___ b__r charts, stacked col___ charts, and tr___ps.
While treemaps are considered 2) sup___r for portraying part-to-whole relationships, here the more commonly used 3) p__e chart is used
1) pie charts, donut charts, stacked bar charts, stacked column charts, and treemaps
2) superior
3) pie
Data Exploration Pattern 6: Correlation
Define Correlation Data Relationship
A data relationship that indicates the 1) de__e to which 2) t__o va___es m__e in the 3) s___e or the 4) op___e dir___n.
For example, if the marketing expenses for a product increase, then it is likely that the sales for the same product also increase
1) degree
2) two variables move
3) same
4) opposite direction.
Data Exploration Pattern 6: Correlation
What are the 2 key features to consider with correlation data relationships ?
-Di___n
-Str___ of the cor___on
-Direction
-Strength of the correlation
Data Exploration Pattern 6: Correlation
2 key features to consider with correlation data relationships
-Direction
-The direction is positive if both variables move in the 1) s__e direction.
-The direction is negative if both variables move in 2) op___e directions
1) same
2) opposite
Data Exploration Pattern 6: Correlation
2 key features to consider with correlation data relationships
-Strength of the Correlation
Strength indicates the degree of 1) co___on between the 2) t__o va__, ranging from 3) n__ correlation to 4) pe___ct correlation
1) correlation
2) tow variables
3) no
4) perfect
Data Exploration Pattern 6: Correlation
The most common visualization for exploring correlation is the 1) sca__t, which is also referred to as a 2) s___er ch__t.
It 3) pl__s the 4) co__es for 5) t___o var__s for each data point.
1) scatterplot
2) scatter chart
3) plots
4) coordinates
5) two variables
Data Exploration Pattern 7: Time Series
Define Time Series Data Relationship
A relationship that defines the values of a variable at 1) seq___ po___s in t___e.
To display a time series, a visualization requires a 2) ti___ unit such as a 3) min___, h___r, d__y, or w__k, and a variable that 4) cha___s o___r t__e.
1) sequential points in time
2) time
3) minute, hour, day, or week
4) changes over time.
Data Exploration Pattern 7: Time Series
Visualizations
A 1) li___ ch___t is the most common visualization for time series, but 2) b__r charts, c___n charts, a__a charts, wa__ll charts, and sp___e charts can also be use
1) line chart
2) bar charts, column charts, area charts, waterfall charts, and sparkline charts
Data Exploration Pattern 7: Time Series
Exploration and Insights
Look for 1) tr___s, cy___s, and irr____s when exploring time series
1) trends, cycles, and irregularities
Data Exploration Pattern 7: Time Series
Exploration and Insights
Look for trends, cycles, and irregularities when exploring time series:
-A trend indicates the 1) ge___ di___n (up___d or dow___) in which a variable moves 2) o__r t__e.
-A cycle indicates a 3) pat___of ch___s, which can vary in 4) le___h and in___ o___r t___e. 5) Se___ty refers to 6) in___a-y___r cy___s with a 7) fixed nature.
-Irregularities are 8) unsy___c, typically sh__t-t___m, flu___n
1) general direction (upward or downward)
2) over time
3) pattern of changes
4) length and intensity over time
5) Seasonality
6) intra-year cycles
7) fixed nature
8) unsystematic, typically short-term, fluctuation
Data Exploration Pattern 8: Geospatial
Define Geospatial Data Relationship
A data relationship in which 1) nu___c values are assigned to 2) lo___s and enc__ by c___r and the s__e of the b___s wi___ the visualization
1) numeric
2) locations and encoded by color and the size of the bubbles within
Data Exploration Pattern 8: Geospatial
Visualizations
Geospatial relationships are defined using maps
-1) Ch___h Map
-2) Pro__al Sy__l Map
1) Choropleth
2) Proportional Symbol
Data Exploration Pattern 8: Geospatial
Visualizations
Geospatial relationships are defined using maps
-Choropleth Map
uses 1) co__ intensity to represent data 2) v___s
1) color intensity
2) values
Data Exploration Pattern 8: Geospatial
Visualizations
Geospatial relationships are defined using maps
-Proportional Symbol Map
uses 1) s___ls—often 2) bu__s/ci__s—and the 3) s__e of the symbol represents the data 4) v__e. The 5) la__r the symbol, the 6) h___r the value.
EX. A business could create a map that shows the total revenue (numeric variable) per city (location variable) for a specific state. The larger the bubble representing the city, the higher the revenue for that city.
1) symbols
2) bubbles/circles
3) size
4) value
5) larger
6) higher
7.3 How are Data Explored by Integrating Data Relationships?
1) Fo____l data relationship with 2) o___e visualization, data exploration often requires integrating 3) t__o or m___ data 4) rel___
5) Int___d data relationships can be represented with a 6) s__le visualization or by using 7) di___t visualizations as part of a 8) rep___.
9) Co__e t__s and 10) P___o an___are two examples of 11) c___d data relationships that can be represented in 12) o__e chart
1) Foundational
2) one
3) two or more
4) relationships
5) Integrated
6) single
7) different visualizations as
8) report
9) Composite trends
10) Pareto analysis
11) combined
12) one
Data Exploration Pattern 9: Composite Trends
Define Composite trend data relationship
An 1) in____d data 2) re___p that shows the 3) changes in a 4) p__t-to-wh__ rel___p over 5) ti___, often 6) an___
1) integrated
2) relationship
3) changes
4) part-to-whole relationship
5) time
6) analyzed
Data Exploration Pattern 9: Composite Trends
The exploration structure for composite trends has three variables and combines the exploration structures of the time series data relationship and part-to-whole data relationship
3 variables
-nominal variable
-numeric variable
-time unit variable
Data Exploration Pattern 9: Composite Trends
Exploration and Insights
Composite trends can be explored for any 1) comb__ of m__s (2) nu___c v___le), 3) dim___s (4) no___al var__), and 5) t__e u__ts by dragging and dropping.
1) combination of measures 2) numeric variable
3) dimensions
4) nominal variable
5) time unit
Data Exploration Pattern 9: Composite Trends
Exploration and Insights
A composite trend pattern can also show how the 1) rel____ im___e of the different 2) ac___s in an 3) inc___ st___nt changed in a 4) sp___c period.
This is a combination of 5) ve___l (6) p__t-t__-wh__e) and 7) ho___l (8) t__e s__s) 9) fin___ st___nt analysis.
1) relative importance
2) accounts
3) income statement
4) specific
5) vertical
6) part-to-whole
7) horizontal
8) time series
9) financial statement
Data Exploration Pattern 10: Pareto Analysis
Define Pareto Analysis
A 1) st___al t___ue that 2) vi___s the 3) im___e of different 4) ca___ies (which is the 5) no___l co__on), 6) ra___s them, and shows how each 7) c___ory con___s to the 8) cum__e p__ge (9) pa__ to w__)
1) statistical technique
2) visualizes
3) importance
4) categories
5) nominal comparison
6) ranks
7) category contributes
8) cumulative percentage
9) part to whole
Data Exploration Pattern 10: Pareto Analysis
Visualizations
A Pareto chart visualizes the combination of the relationships shown in the exploration structure.
Some tools, like Excel, offer Pareto charts.
They can also be created as 1) li___ and co___n charts, which are supported by most tools
1) line and column
Data Exploration Pattern 10: Pareto Analysis
Exploration and Insights
Pareto analysis has a range of applications in business and accounting:
-Identifying the most significant customer complaints. If a small group of issues causes most complaints, removing these issues could positively impact customer satisfaction.
-Indicating main cost categories for a specific project.
-Identifying the employees that generate the most new clients.
-Illustrating which group of holdings in a portfolio is responsible for most of its growth
Reports Using Multiple Visualizations
Data relationships are often integrated using 1) m__e vis__s as part of a 2) re__t.
Mostly 3) int___e, these visualizations provide 4) en___s ex___n op___es.
1) multiple visualizations
2) report
3) interactive
4) endless exploration opportunities.
Reports Using Multiple Visualizations
A report that combines 1) pa__-t__-w__e and ti__e se__ relationships explores 2) tr__s for not only the 3) wh__, but also for each of the 4) in___l parts.
A more advanced example, with multiple interactive visualizations representing 5) dif___ re___s, can illustrate this.
1) part-to-whole and time series relationships
2) trends
3) whole
4) individual
5) different relationships,
Data 1) exp___n patterns are 2) be___ practices for 3) p___ng analysis.
They make us more 4) aw___ of da___ rel__s, how they can be 5) repre___ and exp___, the 6) in___ts that can be 7) ge___ed, and how they can be 8) inte___.
1) exploration
2) best
3) performing
4) aware of data relationships
5) represented and explored
6) insights
7) generated
8) integrated.
8.1 How Do We Draw Conclusions from Data Analysis?
Define Data Analysis Interpretation
The process of 1) ev____ an analysis to 2) und____ and 3) ex___n its 4) me__
the 5) in___s g___d from data analysis interpretation 6) h___p us make 7) g___d bu___s de___s
1) evaluating
2) understand
3) explain
4) meaning
5) insights gained
6) help
7) good business decisions
Data Analysis Interpretation VS. Data Exploration
The goal of data exploration is 1) u____g the d__a, whereas data interpretation involves 2) un___g the a____s.
The 3) f___, however, is what differentiates exploration and interpretation.
1) understanding the data
2) understanding the analysis
3) focus
Data Analysis Interpretation VS. Data Exploration
The first step is exploring the data to 1) un___ it
1) understand
Data Analysis Interpretation VS. Data Exploration
Data analysis interpretation happens at the 1) e___ of the 2) d___a exp____ pr___s.
We move from exploring for insights to 3) inte___g so we can make 4) in___d d___s.
1) end
2) data exploration process
3) interpreting
4) informed decisions
Data Analysis Interpretation
Data analysis interpretation is the process of reviewing an analysis
What are the 2 steps?
Step 1: De___ if the a___s makes s___e
Step 2: V___y that the r___s are v___d and rel___
Step 1: Determine if the analysis makes sense
Step 2: Verify that the results are valid and reliable
Data Analysis Interpretation
-specific questions for each step
Step 1: Determine if the analysis makes sense
Question 1: D__s the a__is a___r the in___d q___on and al___n with the or__al ob___e?
Question 2: Were the c____t data and me____s used to p___m the a___is?
Question 3: Are the re___s r___e, or are more an___s ne___ry?
Question 1: Does the analysis answer the intended question and align with the original objective?
Question 2: Were the correct data and methods used to perform the analysis?
Question 3: Are the results reasonable, or are more analyses necessary?
Data Analysis Interpretation
-specific questions for each step
Step 2: Verify that the results are valid and reliable
Question 4: Does the a___is m___e what it was in___d to m___re?
Question 5: Are the re___s a___e?
Question 4: Does the analysis measure what it was intended to measure?
Question 5: Are the results accurate?
Data Analysis Interpretation
Step 1: Determine if the analysis makes sense
When interpreting our own analysis, the first two questions in Step 1 are answered during the 1) pl___g and a___g stages of the 2) M____C process:
Question 1: Does the analysis answer the original question or objective of the analysis? 3) (Mo___n and Obj___)
Question 2: Were the correct data and appropriate method used to perform the analysis? 4) (St___ and An___s)
1) planning and analyzing
2) MOSIAC
3) (Motivation and Objective)
4) (Strategy and Analysis)
Data Analysis Interpretation
Step 2: Verify that the results are valid and reliable
The remaining questions from both steps would then be answered during the 1) rep___ stage of the data analysis process 2) (Int___n).
However, if someone else prepared the analysis, then it is necessary to address 3) e___h q___n for b___h st__s
1) reporting
2) (Interpretation)
3) each question for both steps
One of the most valued skills of accountants is their ability to be 1) i____ and sk___cal eva___rs of fi___ial info___n.
1) independent and skeptical evaluators of financial information
8.2 What is the Relationship Between Critical Thinking and Data Analysis Interpretation?
What are the 6 elements of critical thinking?
- S___
- P____e
- Alt____s
- Ri___
- Kno___
- Se___-ref____
What’s the acronym ?
- Stakeholders
- Purpose
- Alternatives
- Risks
- Knowledge
- Self-reflection
SPARKS
Stakeholders: Understand the Context
Before 1) int___g the analysis provided by the Celeritas marketing team, 2) co___ who the 3) st___rs are in this 4) de___n.
5) Id___ st___ers (Illustration 8.6) provides 6) i__t into the situation within which the data analysis was 7) cre___. This knowledge can help explain the 8) re___ts.
9) Fa___ to 10) ide___y the 11) sta___s means 12) po___y int____g the 13) re___s from the 14) wr__ perspective.
1) interpreting
2) consider
3) stakeholders
4) decision
5) Identifying stakeholders
6) insight
7) created
8) results
9) Failing
10) identify
11) stakeholders
12) potentially interpreting 13) results
14) wrong
Purpose: Define the “Why” of the Analysis
In addition to recognizing relevant stakeholders, identify the 1) p___e of the analysis.
It is easy to 2) fo____ this and immediately start 3) inte___.
Be careful not to fall into this trap! We 3) c__t fully interpret an analysis until we know its 4) p___e
5) Ove____g the purpose could mean interpreting the analysis 6) inc___
1) purpose
2) forget
3) interpreting
3.1) cannot fully interpret
4) purpose
5) overlooking
6) incorrectly
Alternatives: Investigate Other Interpretations
Always consider if there are 1) alt___ ways to view the 2) re___s of an 3) an___.
Additionally, consider if there are 4) alt___ m__ds to 5) co___ct the analysis that were 6) n__t a__d.
Finally, decide if 7) m__e a___is is 8) ne___
Thinking about these different alternatives helps 9) d___ne whether the most 10) ap___te analysis was used, which can 11) in__e con___e when interpreting the 12) r__ts
1) alternative ways
2) results
3) analysis
4) alternative methods
5) conduct
6) not addressed
7) more analysis
8) needed
9) determine
10) appropriate analysis
11) increase confidence
12) results
Risks: Consider Data, Analysis, and Bias
Examine 1) al__ as__s of the analysis to identify 2) pot___l r__ks. This begins with the data and extends to potential 3) bi__s–b___h o__r o__n and those of 4) sta___s
Asking certain 5) que___s can help evaluate these potential 6) ri___.
Keep in mind that if you are the preparer of the analysis, then you will have 7) alr__y ad__d some of these 8) is___
1) all aspect
2) potential risks
3) biases–both our own
4) stakeholders
5) questions
6) risks
7) already addressed
8) issues
Knowledge: What We Need to Know
To interpret any analysis, determine the 1) kn___ge necessary to 2) un____d it.
We may not have the 3) co__ct ba___ or ex____e, so identifying the 4) re___d un___ng will reveal if we need to do additional 5) re___h
1) knowledge
2) understand
3) correct background or experience
4) required understanding
5) research
Knowledge: What We Need to Know
It helps to ask these questions:
-Is sp___c accounting k___ge nec____?
-Would in___y kn___e be h___ul?
-Is te___y kn___ge important to int___t the analysis?
-Is ad__l re___h necessary, or should we s__k the h___p of an e__t?
-Is specific accounting knowledge necessary?
-Would industry knowledge be helpful?
-Is technology knowledge important to interpret the analysis?
-Is additional research necessary, or should we seek the help of an expert?
Self-Reflection: Think About Lessons Learned
Each data analysis project should include 1) ref__g on le___s learned from pr___s an___s pe___d or int___d.
We can also learn from the 2) c__ent data analysis being 3) pe___d or int___d and apply that to 4) f__re data analysis projects.
Reflecting on 5) pre__ e___nces helps us perform interpretations 6) qu___, th___ly, and ac____y
1) reflecting on lessons learned from previous analyses performed or interpreted
2) current
3) performed or interpreted
4) future
5) previous experiences
6) quickly, thoroughly, and accurately
8.3 How Do We Know the Analysis Makes Sense?
Recall that the first step in data interpretation is determining if the analysis makes 1) se___.
This may seem like an obvious question, but it is one that is often 2) ove____. It is even more important when trying to 3) un____d an analysis of something 4) u____iar.
In terms of data analysis, asking if the analysis “makes sense” means 5) conf____ the analysis has a 6) c___r mea___
1) sense
2) overlooked
3) understand
4) unfamiliar
5) confirming
6) clear meaning
Using 1) cri___l th___g sk___s, determine if the analysis answers the intended 2) qu____n and aligns with 3) ob___, if the 4) co___t d__a and me__ds were used, and if the results are both 5) r____le and su___nt for the project’s 6) pu___.
1) critical thinking skills
2) question
3) objective
4) correct data and methods
5) reasonable and sufficient
6) purpose.
Evaluate the Data and Methods
Questions we can ask to determine if the analysis makes sense include thinking about 1) h___w we got the results:
-Are the d___ used in the analysis re___le given the q___on/ob__e of the analysis?
-Is the a___s me___d reasonable given the qu___/ob___e of the analysis?
1) how
-Are the data used in the analysis reasonable given the question/objective of the analysis?
-Is the analysis method reasonable given the question/objective of the analysis?
Examine the Results
There are also questions that evaluate the reasonableness of the results themselves:
- Are the re___s of the analysis re___e given what is k___n about the s__ct being a___d?
- Are the imp___s of the analysis rea____e given what is known about the sub___ being a___d?
- Does the analysis ad___s the n___s/co___s of the st___rs?
If the answers to any of these questions is “no,” then it is likely that either m___ or a di___nt analysis is nec___y before the r___ts can be int___d.
- Are the results of the analysis reasonable given what is known about the subject being analyzed?
- Are the implications of the analysis reasonable given what is known about the subject being analyzed?
- Does the analysis address the needs/concerns of the stakeholders?
If the answers to any of these questions is “no,” then it is likely that either more or a different analysis is necessary before the results can be interpreted.
Determine if More Information or Analyses are Necessary
Even if the 1) int____n makes 2) s__e, sometimes 3) m___e inf___on or ad___l ana___s are still 4) nec___y to thoroughly answer the 5) que___n.
It is easy to believe we have the 6) inf____n ne___ry to decide on a 7) co__e of ac___n
1) interpretation
2) sense
3) more information or additional analyses
4) necessary
5) question
6) information
7) course of action
Determine if More Information or Analyses are Necessary
This bias can be particularly 1) st___ng if the data analysis 2) su___s 3) prec___d id__s or conc___s about the 4) qu__n.
Accountants must be 5) sk____al evaluators of 6) in___on, so we must 7) dil___ly review data analyses to be sure they provide 8) en__h in___n and su___ort to make a 9) w__l-inf___d dec___n.
1) strong
2) supports
3) preconceived ideas or conclusions 4) question
5) skeptical
6) information
7) diligently
8) enough information and support
9) well-informed decision
Determine if More Information or Analyses are Necessary
What are 2 bias?
-con___n bias
-sel___n bias
-confirmation bias
-selection bias
Determine if More Information or Analyses are Necessary
Define Confirmation Bias
The person 1) perf___g the 2) a__is wants to 3) p__e a 4) p___ned ass___n, so they look for 5) d___a that 6) su___rt their 7) ex___g be__f.
The person 8) int___g the analysis can 9) a___o have this 10) b___s.
Being 10) aw___ of the 11) pr___er’s potential biases, as well as our own, can help to 12) mit___e c___on bias.
1) performing
2) analysis
3) prove
4) predetermined assumption
5) data
6) support
7) existing belief
8) interpreting
9) also
10) bias
10.1) awareness
11) preparer’s
12) mitigate confirmation
Determine if More Information or Analyses are Necessary
Define Selection Bias
This bias occurs when the data 1) u___d in the analysis are 2) s___ed sub___y.
Selection bias is a 3) co___n if the analysis being 4) int____d is based on a 5) sa___e of d___a rather than the 6) en___e po___ion.
An example is when an analysis of sales transactions is based on a sample of transactions rather than on all the transactions. If the sample is 7) n___ a g___d rep___n of the full population, then the results will be 8) bi___d
1) used
2) selected subjectively
3) concern
4) interpreted
5) sample of data
6) entire population
7) not a good representation
8) biased
8.4 How are Validity and Reliability Determined in Descriptive and Diagnostic Analyses?
Once it is clear that the analysis makes sense, the results can be assessed for 1) v___y and rel__lity.
If the results are not 2) v___d, then it does 3) ___t matter how 4) “g__d” the analysis is.
Sometimes it is easy to be fooled by data analyses because we take them at face value
1) validity and reliability
2) valid
3) not
4) “good”
Define Validity
In the context of data analysis, it means that an analysis 1) mea___s what it is 2) su___d to 3) me___e and that it represents 4) re__
1) measures
2) supposed
3) measure
4) reality
Define Reliability
data used are 1) depe___le and tru___y and the 2) m___es used in the 3) an___s are 4) co___nt and ac___e
1) dependable and trustworthy
2) measures
3) analysis
4) consistent and accurate
Summary of Analyses by Analytics Area
The most common types of analyses used in descriptive and diagnostic analytics, how to identify appropriate measures for each type, and how to determine if the results are valid
Descriptive analytics help to better 1) un____ the 2) d__a underlying the analysis being 3) inte___d.
To assess 4) val___y and rel___y of the analyses, ensure the 5) c___ct method is used and that the data are 6) ac___.
Illustration 8.13 is a summary of valid descriptive analyses based on the objective of the analysis.
For an analysis to be 7) v___d, the method 8) m___t m___ch the obj___e. In a 9) re___e analysis, the measures used are 10) ac___e and consistent.
1) understand
2) data
3) interpreted
4) validity and reliability
5) correct method
6) accurate
7) valid
8) must match the objective
9) reliable
10) accurate
Diagnostic Analytics
If the objective of the analysis is to understand 1) w__y so____g h__s oc___d, then you will be interpreting diagnostic analytics
1) why something has occurred,
Diagnostic Analytics
Examine Data Relationships
This measure is a numerical value between 1) __ and __. The higher the absolute number, the 2) gre___r the s___h of the relationship.
A positive correlation coefficient indicates that as one variable 3) inc___s, so does the 4) o___r va___le
A negative correlation is an 5) inv___e relationship. As one variable 6) in___s, the other 7) de___s and vice versa
1) –1 and +1
2) greater the strength
3) increases
4) other variable
5) inverse
6) increases
7) decreases
Data Story Elements
There are three elements to a data story
-d__a, nar___e, and vi___ls
data, narrative, and visuals
Data Story Elements
The author of Effective Data Storytelling, Brent Dykes, describes how these elements combine to 1) ex___n, enli___n, and en__e the audience:
The 2) inte___n of data and 3) nar___e explains the data 4) st__y.
The story’s 5) nar___e provides the 6) co__t and com___ry needed to 7) und___d the 8) re___ts of the 9) an__is.
It provides 10) str___e to the data and 11) gu___s the reader through the 12) me___g of the analysis
1) explain, enlighten, and engage the audience
2) intersection
3) narrative
4) story
5) narrative
6) context and commentary
7) understand
8) results
9) analysis
10) structure
11) guides
12) meaning
Data Story Elements
Data also intersects with 1) vis__s to 2) en__n the reader with 3) ins___ts.
4) Visu___g the data reveals 5) pat___s or t__ds that may have gone 6) un__ed without the help of 7) vi___s.
In fact, humans process visual images 60,000 times faster than text
1) visuals
2) enlighten
3) insights
4) Visualizing
5) patterns or trends
6) unnoticed
7) visualizations
Data Story Elements
Finally, 1) co___g nar___e with 2) vi___s en___es the au___e in the 3) st__y.
A good 4) st___ can hold the 5) at___n of the 6) re___ and 6) inc___s the likelihood of action
1) combining narrative
2) visuals engages the audience
3) story
4) story
5) attention
6) reader
6) increases the likelihood
Data Story Structure
Freytag’s pyramid
It is one of the most 1) t__ht dr___c structures in the world.
1) taught dramatic
Visualization Decision Tree for Showing Composition, Relationships, and Distributions
What is the objective of the analysis ?
-Show co___on
-Show r___hips
-Show di___ons
-Show composition
-Show relationships
-Show distributions
Visualization Decision Tree for Showing Composition, Relationships, and Distributions
What is the objective of the analysis ?
Show composition
-ar__ chart
-p__ chart
-st___d b__r chart
-area chart
-pie chart
-stacked bar chart
Visualization Decision Tree for Showing Composition, Relationships, and Distributions
What is the objective of the analysis ?
Show Relationships
-bu___e chart
-sca____t
-bubble chart
-scatterplot
Visualization Decision Tree for Showing Composition, Relationships, and Distributions
What is the objective of the analysis ?
Show Distributions
-his__m chart
-li__ chart
-sc___lot
-histogram chart
-line chart
-scatterplot
Visualization Decision Tree for Showing Trends or Comparisons
What is the objective of the analysis ?
-Ind__e t___ds
-Com__s
-Indicate trends
-Comparisons
Visualization Decision Tree for Showing Trends or Comparisons
What is the objective of the analysis ?
Indicate Trends
-l__e chart
-co__n chart
-line chart
-column chart
Visualization Decision Tree for Showing Trends or Comparisons
What is the objective of the analysis ?
Comparisons
-It___s
-Ov___e
-Items
-Overtime
Visualization Decision Tree for Showing Trends or Comparisons
What is the objective of the analysis ?
Comparisons
-Items
-b__r chart
-c___n chart
-bar chart
-column chart
Visualization Decision Tree for Showing Trends or Comparisons
What is the objective of the analysis ?
Comparisons
-Overtime
-l__e chart
-co__n chart
-line chart
-column chart
Consider Preattentive Attributes
Color
- Sequential
Color is ordered from l___ to h__
low to high
Consider Preattentive Attributes
Color
- Diverging
There are 1) t__o se__tial c__rs with a 2) ne__al mi___t.
This type of scale is useful for showing 3) ga__s and lo___s
1) two sequential colors
2) neutral midpoint
3) gains and losses
Consider Preattentive Attributes
Color
- Categorical
There are 1) cont__g colors for 2) in___l co___on.
This is a 3) co___on use for color when comparing 4) cat___s
1) contrasting
2) individual comparison
3) common
4) categories
Consider Preattentive Attributes
Position
Item 1) pla___t in 2) vis___s and das___ds matters.
Most people begin viewing a visualization from the 3) t__p l__t c___er and then s___n in a 4) zi__-z___g mo___n through the entire visualization
Therefore, be sure to put the most 5) imp___t inf___on at the 6) t__p le__t of the visualization.
1) placement
2) visualizations and dashboards
3) top left corner and then scan
4) zig-zag motion
5) important information 6) top left
Before and After Reducing Visualization Clutter Chart
Cluttered Visualization
-T___ many co__rs are n__t adding to the m__ng of the gr__h because the models are also dir__y la___d.
Reduced Visualization
-The vi___l is a si__e co__r.
Cluttered Visualization
-Too many colors are not adding to the meaning of the graph because the models are also directly labeled.
Reduced Visualization
-The visual is a single color.
Before and After Reducing Visualization Clutter Chart
Cluttered Visualization
-A__s l__ls are provided, but the audience m__t g__s the ex__t a___nt.
Reduced Visualization
-La__l e__h b__r dir___y with the sales volume am___t and rem__ the a__s.
-Sales volume is in the ti__e of the graph, so r___ve the a__s he__r.
Cluttered Visualization
-Axis labels are provided, but the audience must guess the exact amount.
Reduced Visualization
-Label each bar directly with the sales volume amount and remove the axis.
-Sales volume is in the title of the graph, so remove the axis header
Before and After Reducing Visualization Clutter Chart
Cluttered Visualization
-An___n t___t b___es identify the bes___g model
Reduced Visualization
-S__t the d__a from hi___st to lo___t
Cluttered Visualization
-Annotation text boxes identify the bestselling model
Reduced Visualization
-Sort the data from highest to lowest
Before and After Reducing Visualization Clutter Charter
Cluttered Visualization
-Bac____d c___r of the graph is un___ary.
Reduced Visualization
-R___ve bac___d co__r.
-Re___e g___es since we are rem___g the a__is and lab___ng the col__s directly.
Cluttered Visualization
-Background color of the graph is unnecessary.
Reduced Visualization
-Remove background color. Remove gridlines since we are removing the axis and labeling the columns directly.
Before and After Reducing Visualization Clutter Charter
Cluttered Visualization
-Models are listed in alp___al order. There is n__ ind___on which model the audience should fo__s on.
Reduced Visualization
-The bestselling model is hig___d and the others are gr__d o__t to draw at__ion to it.
Cluttered Visualization
-Models are listed in alphabetical order. There is no indication which model the audience should focus on.
Reduced Visualization
-The bestselling model is highlighted and the others are grayed out to draw attention to it.
Use Visualization-Specific Best Practices
Area Chart
-Do n__t use with more than fo___ cat___es to avoid con___n and clu__er.
-Start the _-a__s at ze__o.
-Put h___y va___e data on the t___p and data with l__w va__ty on the bo___m.
-Do not use with more than four categories to avoid confusion and clutter.
-Start the y-axis at zero.
-Put highly variable data on the top and data with low variability on the bottom.
Use Visualization-Specific Best Practices
Bar and Column Chart
-Use ho___l bars if there are more than __ categories or l___g category la___ls.
-Use ho___al la__ls for better rea__ty.
-Sp___e bars app__ely and con___ly.
-Use co___r s___gly, or as an ac__nt.
-Always have a z__o baseline (the y-a__s begins at ze___).
-Compare __–__ categories with ve___l columns.
-Use horizontal bars if there are more than 7 categories or long category labels.
-Use horizontal labels for better readability.
-Space bars appropriately and consistently.
-Use color sparingly, or as an accent.
-Always have a zero baseline (the y-axis begins at zero).
-Compare 2–7 categories with vertical columns.
Use Visualization-Specific Best Practices
Column Chart
-La__l bu___es and make sure they are vi___e.
-Sc___e bubble s__e by a__a and not di__r.
-Do n__t use bubbles if they are all si___r in si__e.
-Label bubbles and make sure they are visible.
-Scale bubble size by area and not diameter.
-Do not use bubbles if they are all similar in size.
Use Visualization-Specific Best Practices
Histogram Chart
-Use a z__o b___ne.
Choose an ap___te n___er of bi__s:
-Bins are nu___rs that represent the in__ls into which the data will be gr___ed.
-Bins define the gro___s used for the freq___y dis___ution.
-Generally, include between ___–__ bins.
-Use a zero baseline.
Choose an appropriate number of bins:
-Bins are numbers that represent the intervals into which the data will be grouped.
-Bins define the groups used for the frequency distribution.
Generally, include between 5–15 bins.
Use Visualization-Specific Best Practices
Line Graph
-Ti___ runs from l__t to r___ht.
Be con___t when plotting t__e poi__s.
-Use so__d li__s, not d__d.
-Use a z__o ba___ne.
-Do n__t p__t more than f__ur lines. Use mu__e ch___ts, instead.
-Time runs from left to right.
Be consistent when plotting time points.
-Use solid lines, not dotted.Use a zero baseline.
-Do not plot more than four lines. Use multiple charts, instead.
Use Visualization-Specific Best Practices
Pie Chart
-Most im___ul with s__ll data sets.
-Best to use when showing dif___s wit___n gr__s based on o__e variable.
-Ensure the data adds to __%.
L__it the chart to a ma___um of fi___ s___ts.
-Start the fi__t segment at __ o’clock position
-Most impactful with small data sets.
-Best to use when showing differences within groups based on one variable.
-Ensure the data adds to 100%.
Limit the chart to a maximum of five segments.
-Start the first segment at 12 o’clock position
Use Visualization-Specific Best Practices
Stacked Bar Chart
-Can be v__al or ho___ntal.
-Follow s__e b__t practices as b__r ch__ts.
-Used to show c___ns of su___ents between cat___s.
-Best used when there are n__t too m__y su__ents.
-Consider using a __% stacked bar to make co___ons between the b__s and sub__nents easier
-Can be vertical or horizontal.
-Follow same best practices as bar charts.
-Used to show comparisons of subcomponents between categories.
-Best used when there are not too many subcomponents.
-Consider using a 100% stacked bar to make comparisons between the bars and subcomponents easier
Use Visualization-Specific Best Practices
Scatter Chart
-Data set should be in p__rs with an ind___nt variable (__-axis) and a de__t variable (__-axis).
-Use if or__r is n__t re___t–otherwise use a l__e graph.
-Do n__t u__e if there are only a f___w pi___s of data or if there is n__ cor___tion.
-Data set should be in pairs with an independent variable (x-axis) and a dependent variable (y-axis).
-Use if order is not relevant–otherwise use a line graph.
-Do not use if there are only a few pieces of data or if there is no correlation.
Use Visualization-Specific Best Practices
Tree Map
-App___te when p__e com__ns are n__t important.
-Use br__ht, co__g colors so that each box is easily de__ned.
-L___el boxes with t__t or nu__rs.
-Appropriate when precise comparisons are not important.
-Use bright, contrasting colors so that each box is easily defined.
-Label boxes with text or numbers.
Manipulating the Y-Axis
Like omitting the baseline, 1) ma__ng the 2) sc__e on the 3) __-axis can 4) af___t how the data are 5) int__ed.
6) Exp___g or co___ng the 7) scale on the 8) __-axis can make 9) ch___es in the data seem 10) m__e or le__s sign__nt.
1) manipulating
2) scale
3) y-axis
4) affect
5) interpreted
6) Expanding or compressing 7) scale
8) y-axis
9) changes
10) more or less significant.
Going Against Conventions
A general standard in data visualization is that 1) da__r colors indicate 2) hi__r numbers in a color scale
1) darker
2) higher
Selectively Choosing the Data
Including only 1) s__e data po__s in a visualization may also create a 2) fa___e im___n of the data
1) some data points
2) false impression
Using the Wrong Type of Graph
Sometimes choosing the 1) wr__g ty__e of gr__h can make it 2) di___lt to 3) inte___t the data and 4) res__s in 5) mi___ng the viewer.
A visualization that is 6). n__t ap___ate for the 7) t___e of d__a or an___sis res___ts being re___ed makes it 8) dif___lt for the audience to 9) in___t the 10) me__e.
1) wrong type of graph
2) difficult
3) interpret
4) results
5) misleading
6) not appropriate
7) type of data or analysis results being reported
8) difficult
9) interpret
10) message
