Questions & explanations
1. Give an example of a system that is controllable but not observable.
Consider a system with two states: A = [[0,1],[0,0]], B = [[0],[1]], and C = [1,0]. The system is controllable because the controllability matrix [[0,1],[1,0]] has rank 2. But the observability matrix is [[1,0],[0,1]] which also has rank 2, so actually observable. Need different C. Try C = [1,1]. Then observability matrix = [[1,1],[0,1]] has rank 2, still observable. For uncontrollable but observable case: A = [[1,0],[0,2]], B = [[1],[0]], C = [[1,0]]. Controllability matrix = [[1,1],[0,0]] rank 1, not controllable. Observability matrix = [[1,0],[1,0]] rank 1, not observable. Actually that example is neither. A classic controllable but not observable: two plants in parallel with same input but output only one. Example: A = diag(1,2), B = [1;1], C = [1,0]. Controllability matrix = [[1,1],[1,2]] rank 2 controllable. Observability = [[1,0],[1,0]] rank 1, not observable. So the second state (eigenvalue 2) is not seen in output.
2. Design a simple rubric criterion that assesses critical thinking without cultural bias.
A fair criterion is: 'The work shows how the student connects ideas from different sources or experiences – the connection can be any form: comparison, cause-effect, or application to a personal context.' This allows students from any culture to demonstrate thinking in their own way. It does not require a specific structure like 'claim-evidence-warrant.' It also accepts personal anecdotes if relevant. The descriptor for top performance could be: 'Makes insightful connections that deepen understanding.' This criterion focuses on the process of linking, not on a particular cultural style of argument. It is open to multiple ways of showing intelligence.
3. Why might a checklist of skills be seen as too narrow in a non-Western educational context?
A checklist of skills assumes that learning can be divided into small, separate parts. Many non-Western cultures see learning as integrated and social. For example, in some African philosophies, knowledge is communal and includes wisdom, not just skills. A checklist might miss important qualities like respect, generosity, or community service. Also, checklists tend to focus on easily measured abilities, ignoring character and relationships. In these contexts, a broader assessment that includes contributions to the group and personal growth is more valued. Thus, checklists feel too narrow and may not reflect what the community values in education.
4. How can a teacher use a rubric to prevent plagiarism before students start writing?
Before students begin, the teacher can show the rubric and discuss what each criterion means. For instance, if the rubric says 'uses at least three sources and cites them correctly,' students know they must find and credit sources. The teacher can also include a criterion like 'shows original analysis of the topic.' This pushes students to think for themselves rather than copy. By making the rubric public, students understand that copying will be easy to catch because their work must match specific, original content expectations. The teacher can also give examples of poor and good work based on the rubric, teaching what plagiarism looks like.
5. How does a holistic assessment differ from a rubric in evaluating student work?
A holistic assessment looks at the whole work and gives one overall judgment, like 'excellent' or 'needs improvement,' based on the teacher's overall impression. A rubric breaks the work into separate criteria (e.g., content, organization) and scores each. Holistic assessment is faster but can be less detailed. In some non-Western cultures, holistic assessment is preferred because it sees the student as a whole person and values harmony between parts. For example, a teacher might say 'this project shows deep understanding and effort' without rating each piece. This aligns with educational philosophies that avoid reducing learning to numbers.
6. Compare LQR with pole placement control. How are they different?
Both use state feedback to place closed-loop poles, but the method differs. In pole placement, you directly choose desired pole locations and solve for K that places poles exactly there. In LQR, you do not choose poles explicitly; instead, you choose Q and R weights, and the algorithm finds the optimal K that minimizes the cost. LQR automatically places poles at locations that give a trade-off between performance and control effort. Pole placement can be easier for meeting specific time-domain specs (like settling time), but LQR provides optimality and robustness (good gain and phase margins). LQR is more systematic for multi-input systems.
7. Compare a phase-locked loop with a simple feedback oscillator. How are they different?
A simple feedback oscillator uses the Barkhausen criterion to generate a fixed frequency signal without an input reference. A PLL, in contrast, requires an input reference signal and its output synchronizes to that reference. The oscillator in a PLL is a VCO whose frequency is controlled by a voltage, while in a simple oscillator the frequency is fixed by passive components. The feedback in an oscillator maintains amplitude and phase conditions for self-sustained oscillation; in a PLL, feedback actively drives the output to match an external signal. PLLs can lock to a range of frequencies, whereas oscillators generate a single frequency.
8. Give an example of how a rubric can detect dishonesty in a student's project.
Suppose a rubric for a science report includes a criterion: 'explains the experiment's steps in your own words.' If a student writes steps that use unusual terms from an online source without changing them, the teacher can notice. The rubric expects personal rewriting, so copying shows as a mismatch. Another example: if the rubric asks for a personal opinion on results, and the student gives a generic statement found on the web, that is a red flag. Teachers can compare the rubric's language to the student's work to spot outsourced content. Rubrics also help by making the expected effort clear, so students who take shortcuts stand out.
9. Why is it important that rubrics for medical skills are tested for reliability?
Reliability means that different examiners would give the same student the same score. In medical education, if two examiners watch the same OSCE station, they should agree. If a rubric is vague, like 'good communication,' examiners might judge differently. A well-designed rubric with clear, behavioral descriptors (e.g., 'makes eye contact, uses a calm voice') increases reliability. This is crucial because medical certification decisions affect patient safety. Unreliable rubrics could let a weak student pass or a good student fail unfairly. So, medical schools pilot-test their rubrics and train examiners to use them consistently.
10. Explain how systems thinking can improve patient discharge processes.
Systems thinking looks at the whole process, not just one step. For discharge, issues often include delayed medication reconciliation or transport. A leader maps the steps from discharge order to patient leaving. They find bottlenecks, like pharmacy delays. Then they redesign: have a discharge pharmacist verify meds early. They also create a checklist for nurses. After changes, they measure time from order to discharge. For example, reducing wait time by 30 minutes. This approach involves all departments. The leader must communicate with each team to ensure smooth change. Systems thinking improves efficiency and reduces errors.
11. What is a phase-locked loop (PLL)?
A phase-locked loop is a feedback system that synchronizes the phase and frequency of an output signal to a reference signal. It consists of three main parts: a phase detector, a loop filter, and a voltage-controlled oscillator (VCO). The phase detector compares the phase of the reference signal and the feedback from the VCO. The difference creates an error voltage that is filtered and then adjusts the VCO frequency to minimize the phase difference. When locked, the output frequency equals the reference frequency and the phase difference is constant. PLLs are used in FM demodulation, frequency synthesis, and clock recovery.
12. Why is stability analysis important in MRAC, and what can go wrong?
Because the controller parameters change over time, the closed-loop system might become unstable if the adaptation is too fast or the wrong rule is used. Stability analysis ensures that the adaptation law will not make the system diverge. For example, the MIT rule can be simple but may not guarantee stability; Lyapunov-based methods provide stability guarantees. Poorly designed MRAC can lead to parameter drift, high-frequency oscillations, or even loss of control. Therefore, careful choice of the reference model and adaptation gains is crucial. Adaptive laws must include leakage or projection to keep parameters bounded.