Free Practice Questions for WGU Foundations-of-Computer-Science Exam

In NumPy, every ndarray has a shape attribute that describes the size of the array along each dimension. For a 2D array, shape returns a tuple with two integers: (number_of_rows, number_of_columns). For example, if a = np.array([[1, 2, 3], [4, 5, 6]]), then a.shape is (2, 3), meaning 2 rows and 3 columns. This is a fundamental idea in matrix and array computing, because shape governs how indexing, slicing, broadcasting, and linear algebra operations behave.

Option A describes the dtype, which can be accessed with a.dtype, not a.shape. Option C is incorrect because shape provides per-dimension sizes, not their sum. Option D refers to the total number of elements, which NumPy provides via a.size (or equivalently np.prod(a.shape)).

Textbooks emphasize shape because many errors in numerical computing come from mismatched dimensions. For example, matrix multiplication requires compatible inner dimensions, and broadcasting rules depend on dimension sizes. By checking .shape, programmers can verify their data layout before applying algorithms, ensuring rows represent observations and columns represent features (or vice versa). Thus, for a 2D NumPy array, .shape indicates the number of rows and columns.

In Python, slicing is the standard mechanism for extracting arangeof elements from a sequence type such as a list, string, or tuple. The character that signals a slice range is thecolon:. The general slice syntax is sequence[start:stop:step]. Most commonly, you see sequence[start:stop], where start is the index to begin from (inclusive) and stop is the index to end at (exclusive). This “inclusive start, exclusive stop” rule is emphasized in textbooks because it makes slice lengths easy to reason about: when step is 1, the number of elements returned is stop - start.

For example, if items = ["a", "b", "c", "d", "e"], then items[1:4] returns ["b", "c", "d"]. Omitting start defaults to the beginning (items[:3] gives the first three elements), and omitting stop defaults to the end (items[2:] gives everything from index 2 onward). The optional step supports patterns like items[::2] for every other element, and negative steps can reverse a sequence (items[::-1]).

The other characters do not define ranges in Python slicing: , separates items (or indices in multidimensional structures), + is addition/concatenation, and = is assignment. The colon is the slicing operator that indicates a range.

NumPy slicing uses the same start/stop rules as Python sequences, and it also supports negative indices to count from the end. In a 2D array, slicing is written as array[rows, columns]. To get thelast two rows, you use -2: in the row position, meaning “start two rows from the end and go to the end.” Similarly, to get thelast two columns, you use -2: in the column position. Combining these gives array[-2:, -2:], which selects the bottom-right 2×2 subarray.

Option A, array[-2:, :], selects the last two rows butall columns, so it is not restricted to the last two columns. Option D, array[:, -2:], selects all rows but only the last two columns. Option B, array[-1:, -1:], selects only the last row and the last column, producing a 1×1 (or 1×1 view) subarray, not a 2×2.

This kind of slicing is widely taught because it is essential for matrix operations, extracting submatrices, working with sliding windows, and manipulating image or time-series data where “take the last k observations/features” is common. Negative indexing reduces errors and makes code clearer, especially compared with computing explicit indices like array[rows-2:rows, cols-2:cols].

AnAcceptable Use Policy (AUP)defines how employees and users are permitted to use an organization’s computing resources—such as email, internet access, file storage, endpoints, and networks—and it typically specifies prohibited behaviors and the consequences of violations. In security and IT governance textbooks, the AUP is framed as both a behavioral contract and a risk-management tool: it reduces misuse, clarifies expectations, and provides an enforceable basis for disciplinary action.

The “ramifications of abusing company resources” (for example, installing unauthorized software, excessive personal use, accessing inappropriate content, attempting to bypass security controls, or sharing credentials) are precisely the kinds of issues an AUP addresses. The policy often includes monitoring statements (users have limited expectation of privacy), compliance requirements, and escalation paths for violations.

A Network Security Policy (A) focuses on technical rules for network protection—firewalls, segmentation, remote access, and intrusion detection—rather than broad user conduct and disciplinary consequences. A Physical Security Policy (B) addresses protection of facilities and hardware—badges, locks, visitor procedures, secure areas. A Data Retention Policy (D) defines how long data is stored, how it is archived, and how it is disposed, which is different from defining misuse consequences.

Thus, the policy aspect that defines permissible behavior and the consequences for abusing resources is the Acceptable Use Policy.

Outlier detection often relies on measuring how far values deviate from a “typical” center. While variance and standard deviation can be used in simple z-score based methods, they arenot robust: a few extreme outliers can inflate the mean and standard deviation, masking the very outliers you want to find. A widely taught robust alternative is themedian absolute deviation (MAD), which is based on the median rather than the mean and therefore resists distortion by extreme values.

MAD is computed by first taking the median of the data, then computing the absolute deviation of each point from that median, and finally taking the median of those deviations. Because medians are stable under extreme values, MAD provides a strong baseline for identifying unusually distant points. Many textbooks and data analysis references present MAD as a robust scale estimator for outlier detection, often combined with a threshold rule such as flagging points whose deviation exceeds a constant multiple of MAD (with a scaling factor sometimes used to make it comparable to standard deviation under normality assumptions).

In NumPy, you can implement MAD using np.median() and np.abs(). Mode is generally not useful for continuous numeric outlier detection, and variance/standard deviation are more sensitive to outliers than MAD. Thus, among the given options, the best statistical measure for detecting outliers robustly is the median absolute deviation.

When NumPy constructs an ndarray, it chooses a single data type called the dtype for the entire array. This is a defining feature of NumPy arrays: unlike Python lists, which can hold mixed object types freely, a NumPy array is designed for efficient numerical computation by storing values in a uniform, contiguous representation. Therefore, if you provide mixed types at creation time, NumPy will select a dtype that can represent all provided values and will convert elements as needed.

This process is commonly described as type promotion or coercion to a common type. For example, mixing integers and floats produces a float array because floats can represent integers without loss of generality. Mixing numbers and strings often results in a string dtype (or, in some cases, an object dtype), because numbers can be converted to their string representations. Once the dtype is chosen, the array behaves consistently under vectorized operations appropriate for that dtype.

Option B correctly summarizes this textbook behavior: the array will contain a single type, converting all elements to that type. Option A is too absolute—many mixed-type arrays still support calculations depending on the resulting dtype. Option C is vague and misses the crucial fact that conversion occurs. Option D is not how NumPy works; it never automatically splits inputs into multiple arrays by type.

Understanding dtype coercion matters because it affects memory usage, performance, and whether numerical operations behave as expected.

Factorial and Fibonacci are classic examples used to teachrecursion, a technique where a function solves a problem by calling itself on smaller subproblems. The key requirement for recursion is (1) abase casethat stops further calls and (2) arecursive casethat reduces the problem size. For factorial, the definition is (n! = n \times (n-1)!) with base case (0! = 1) (or (1! = 1)). For Fibonacci, (F(n) = F(n-1) + F(n-2)) with base cases (F(0)=0) and (F(1)=1). These mathematical definitions map directly into recursive code, which is why textbooks frequently introduce recursion using these sequences.

While factorial and Fibonacci can also be computed iteratively, the question asks for the principle that can be used to implement such algorithms, and recursion is the canonical textbook answer. Recursion also connects to important CS topics: call stacks, activation records, and divide-and-conquer problem solving.

Option A (“procedural programming”) and option D (“object-oriented programming”) are broader paradigms rather than the specific technique used in the classic implementations. Option B (“iterative programming”) is a valid alternative approach, but the standard instructional principle highlighted for these particular examples is recursion. Textbooks also note that naive recursive Fibonacci is inefficient (exponential time) unless optimized with memoization or converted to an iterative or dynamic programming approach.

NumPy supports a powerful technique calledBoolean indexing(also called Boolean masking) to filter arrays based on a condition. When you write bmi > 23, NumPy performs an element-wise comparison and produces a Boolean array of the same shape, containing True where the condition holds and False otherwise. Using that Boolean array inside square brackets, as in bmi[bmi > 23], tells NumPy to return a new 1D array containing only the elements whose mask value is True. This approach is heavily emphasized in scientific computing curricula because it expresses selection logic without explicit loops and runs efficiently in optimized compiled code.

Option B looks close but is not standard NumPy usage. The function commonly used is np.where(condition) or np.where(condition, x, y). While np.where(bmi > 23) can return indices, bmi.where(...) is not a NumPy array method; it is more associated with pandas objects. Options A and C are not valid NumPy APIs for filtering.

Boolean indexing is central in data analysis tasks such as removing invalid measurements, selecting a population subgroup, applying thresholds, and building feature subsets. It composes cleanly with vectorized computation, for example bmi[bmi > 23].mean(), enabling concise and high-performance numerical workflows.

When a Bluetooth device does not configure automatically, the underlying issue is often driver discovery, device enumeration, or the Bluetooth adapter’s state. In Windows, the tool traditionally associated with manually managing hardware devices and their drivers isDevice Manager. It lets a user view hardware categories (including Bluetooth adapters), enable or disable devices, update drivers, uninstall and rescan, and address “unknown device” situations. These actions are core to manual configuration because they influence whether Windows can properly recognize and communicate with a Bluetooth device.

Windows 11 pairing itself is typically initiated from the Settings app under Bluetooth and devices, where a user chooses “Add device” to pair a new accessory. (Microsoft Support) However, among the options provided, only Device Manager is a hardware-configuration tool that can resolve situations where automatic configuration fails due to driver or adapter problems. Network-related tools do not handle local device drivers, Task Scheduler automates tasks rather than adding devices, and Windows Defender is focused on security and malware protection rather than device setup.

From a systems perspective, this reflects a key operating-systems concept: successful device use requires both discovery/pairing and a correctly installed driver stack. Device Manager is the standard interface for the driver and device side of that equation, which is why it is the best match to “manually add or configure” hardware in the given choices.