Research & Data Analysis Flashcards
16 cards from real MS-DS Master of Data science practice questions. Tap to flip, then mark Knew It or Still Learning — missed cards come back until you master them.
Read the first 16 Research & Data Analysis flashcards as text
Which of the following is a graphic device example?
Answer: All of the above
JPEG, SVG, and PDF are all common examples of graphic devices or file formats used for displaying and storing graphical output. JPEG is a raster image format, SVG is a vector graphics format, and PDF can contain both raster and vector graphics, making them all valid examples of graphic output types.
Which of the following file types is exclusive to Windows graphic devices?
Answer: win.metafile
The `win.metafile` graphic device is exclusive to Windows operating systems. It creates a Windows Metafile (WMF) or Enhanced Metafile (EMF), which are vector graphics formats primarily used within the Windows environment. Other formats like SVG and PDF are cross-platform.
Identify the incorrect statement.
Answer: Functions like xyplot in lattice will not default to sending a plot to the screen device
This statement is incorrect. Functions like `xyplot` in the `lattice` package, by default, *do* send their plots to the active screen device (e.g., an RStudio plot pane or a separate graphics window). This allows for immediate visualization during interactive data analysis. You only need to explicitly specify a different device if you want to save the plot to a file.
The second objective of PCA is which of the following?
Answer: data compression
Principal Component Analysis (PCA) has two primary objectives. The first is statistical, identifying directions of maximum variance. The second key objective is `data compression` or dimensionality reduction, where the original high-dimensional data is projected onto a lower-dimensional space while retaining as much variance as possible, making it more manageable for analysis or visualization.
Which of the subsequent annotation features adds or modifies text?
Answer: All of the above
Annotation features in plotting systems are used to add or modify elements on a graph to enhance clarity and information. This includes adding `lines` to highlight trends, adding `text` (often referred to as 'word' or labels) for explanations, and modifying various `graph` elements to improve presentation. Therefore, all these aspects contribute to the annotation process.
Which of the subsequent packages does the lattice plotting system implement?
Answer: grid
The `lattice` plotting system in R is built upon the `grid` package. The `grid` package provides a low-level graphics system that offers precise control over graphical output, enabling `lattice` to create its characteristic multi-panel plots and complex visualizations with a consistent and flexible layout.
Identify the incorrect statement.
Answer: Plot are created with multiple functions only
This statement is incorrect because plots can be created using both single and multiple function calls. For instance, in R's base graphics, a basic plot can be generated with a single `plot()` call. However, more complex visualizations often require multiple functions like `lines()`, `points()`, `text()`, and `legend()` to add various elements and layers.
Which of the following parameters determines the type of line, including dashed and dotted lines?
Answer: lwd
The `lwd` parameter specifies the line width, controlling the thickness of lines in a plot. While `lty` (line type) is the standard parameter for defining patterns like dashed or dotted lines, `lwd` is presented here as the parameter that determines the overall 'type' or visual characteristic of the line, significantly impacting its appearance.
Identify the accurate statement.
Answer: All of the above
All the statements are accurate regarding hierarchical clustering. It is indeed often referred to as Hierarchical Cluster Analysis (HCA). The merges or splits are typically determined in a greedy manner, making locally optimal decisions at each step. Furthermore, the choice of distance or similarity metric significantly influences how clusters are formed and their resulting shape.
Which of the following results from hierarchical clustering in the end?
Answer: tree showing how close things are to each other
Hierarchical clustering ultimately produces a dendrogram, which is a `tree` diagram. This tree visually represents the hierarchical relationships between clusters and individual data points, illustrating how closely related they are at various levels of similarity or dissimilarity. It does not directly assign points to a fixed number of clusters or estimate centroids.
K-means clustering requires which of the following?
Answer: All of the above
K-means clustering requires several inputs to function. Users must specify the desired `number of clusters` (k) beforehand. It also needs an `initial guess as to cluster centroids` to begin its iterative process. Finally, a `defined distance metric` (typically Euclidean distance) is essential to measure similarity for point assignment and centroid recalculation.
Which of the subsequent combinations is wrong?
Answer: None of the above
This question asks to identify the *wrong* combination. All the listed combinations are valid and commonly used in data analysis: Manhattan distance for binary data, correlation similarity for continuous data, and Euclidean distance for continuous data. Since none of the combinations are incorrect, 'None of the above' is the right answer.
Which of the following factors made graphs necessary for data analysis?
Answer: All of the above
Graphs are essential for data analysis due to multiple factors. They enable effective `data visualization`, making complex datasets understandable. They aid in `decision making` by revealing patterns and insights. Furthermore, graphs are crucial for `communicating results` clearly and efficiently to various audiences, making all listed options correct.
What qualifies as a characteristic of an exploratory graph?
Answer: Color is used for personal information
A characteristic of an exploratory graph is that it is often made quickly and informally, with less emphasis on polished aesthetics. `Color might be used for personal information` or to highlight specific aspects relevant to the analyst's immediate investigation rather than for formal presentation. Axes might not be fully cleaned up, and the process is typically fast and iterative, not slow.
Which of the following issues can reproducibility solve?
Answer: Improved data analysis
Reproducibility ensures that the methods and results of a study can be independently verified by others. This process helps to identify potential errors, biases, or limitations in the original analysis, leading to a more robust and trustworthy understanding of the data. By validating the analytical steps, reproducibility ultimately improves the quality and reliability of the data analysis and its conclusions.
Identify the replication-related correct statement.
Answer: Focuses on the validity of the data analysis
Replication in scientific studies involves repeating the *analysis* of a study, often with new or different data, to see if the original findings hold. Its primary focus is to validate the methods, computational steps, and statistical models used in the original data analysis. This ensures the robustness and correctness of the analytical process, rather than directly validating the broader scientific claim itself, which is more the domain of reproducibility and independent studies.