Data visualization is the first step in communicating and displaying the results of data analysis. Commonly known data visualization techniques are tables and charts. Visualization techniques are selected based on the information that must be conveyed to the users. In your discussion,
Categorize situations where tables are a preferred visualization technique. Illustrate with an example.
Choose five different charting techniques that are available for visualization and compare. Explain why some are not suitable for displaying data and why others are more effective for data displays.
Explain Simpson’s paradox and provide an example different from the one listed in your textbook.
Sample Solution
Tables are a preferred visualization technique for displaying data in situations where it needs to be presented in an organized, concise manner. Tables are most often used when the data has several variables or attributes that can easily be compared and contrasted with each other. For example, a table may be used to show average monthly salary ranges in different cities along with the population size of each city and/or differences between wages earned by men and women.
Sample Solution
Tables are a preferred visualization technique for displaying data in situations where it needs to be presented in an organized, concise manner. Tables are most often used when the data has several variables or attributes that can easily be compared and contrasted with each other. For example, a table may be used to show average monthly salary ranges in different cities along with the population size of each city and/or differences between wages earned by men and women.
Five common charting techniques available for visualizing data include line graphs, bar graphs, scatter plots, histograms, and pie charts. Line graphs are useful for displaying trends over time while bar graphs compare values across categories; however these techniques may not necessarily convey the full story if they do not take into consideration all possible factors involved (i.e. outliers). Scatter plots can display relationships between two sets of values but may not reveal any hidden patterns within larger datasets; thus making them less suitable for complex analysis tasks. Histograms on the other hand focus on showing distributions of numeric data which makes them well-suited for spotting outliers that line or bar charts would miss. Lastly pie charts provide good visuals for comparing parts of a whole such as relative percentages among different groups but should only be used when there are no more than five categories being compared since it can become increasingly difficult to interpret when more segments are added.
Simpson’s paradox is a phenomenon that occurs when two seemingly unrelated datasets show contradictory results; usually due to some kind of bias or confounding variable in one of them which is unaccounted for during analysis. A classic example is gender discrimination: An employer might appear to have lower rates of hiring women than men overall but if you break it down into individual departments you might uncover some interesting trends – e..g., although fewer women applied overall they were accepted at higher rates than their male counterparts when applying directly within certain departments like Marketing or Human Resources.
Another example could involve hospital admissions – say you look at the rate at which people admitted with heart conditions survive after one year but then further analyze this by age group you might find that older people seem much less likely to survive even though younger people had better outcomes overall; this could potentially be indicative of incorrect diagnosis taking place due to ageism playing a role during initial evaluations before admission was granted
