To find the probability that all selected transistors are defective, we use the hypergeometric distribution formula:
\[
P(X = k) = \frac{\binom{K}{k} \binom{N-K}{n-k}}{\binom{N}{n}}
\]
Substituting the values \(N = 16\), \(K = 4\), \(n = 4\), and \(k = 4\):
\[
P(X = 4) = \frac{\binom{4}{4} \binom{12}{0}}{\binom{16}{4}} = \frac{1 \cdot 1}{1820} = 0.000549
\]
Thus, the probability that all are defective is approximately \(0.0005\).
Next, we calculate the probability that none of the selected transistors are defective. Again, we use the hypergeometric distribution formula:
\[
P(X = k) = \frac{\binom{K}{k} \binom{N-K}{n-k}}{\binom{N}{n}}
\]
Substituting the values \(N = 16\), \(K = 4\), \(n = 4\), and \(k = 0\):
\[
P(X = 0) = \frac{\binom{4}{0} \binom{12}{4}}{\binom{16}{4}} = \frac{1 \cdot 495}{1820} = 0.272
\]
Thus, the probability that none are defective is approximately \(0.272\).
The probabilities are:
- Probability that all are defective: \( \boxed{0.0005} \)
- Probability that none are defective: \( \boxed{0.272} \)