Bounds, counting and linear structure
Read expected values in Book Isle: 1.75 need not be a possible single score
Weight scores 0, 2, 5 correctly to obtain 7/4, distinguish a reciprocal error and verify a linear score transformation.
On this page
An expected score 1.75 does not predict that score on the next trial. Use an original nonmonetary score card in Book Isle to align values, probabilities and weighted products, separating expectation, mode and observed sample mean. Numbers are stipulated, no draw was performed and no statistical engine is assumed.
List the whole distribution rather than three scores alone
E1 gives X=0, 2, 5 with probabilities 1/2, 1/4, 1/4. Outcomes are mutually exclusive and exhaustive, probabilities nonnegative and summing to one. Align each value and probability, import a clear PDF/EPUB and bookmark E1 after checking rows. Three values do not imply equal chances. Add any missing outcomes from an encountered table before applying this three-row example. Units are fictional exercise points, not real returns.
Multiply each value by its own probability
Compute 0×1/2=0, 2×1/4=1/2 and 5×1/4=5/4, not the deliberately proposed 4/5. Their sum is 0+2/4+5/4=7/4=1.75. Probability weights a score; do not divide by it. Also retain reciprocal error 4/7 as a control, since expectation does not invert the result. Products have score units while probabilities are dimensionless.
x P(X=x) x*P(X=x)
0 1/2 0
2 1/4 1/2
5 1/4 5/4
sum 1 E[X]=7/4=1.75
Check using four equally likely selector positions
Represent the model by four equal-chance positions 0, 0, 2, 5. Their mean 7/4 agrees with weighting. Two zeros give one score two probability shares, not a fourth distinct score. This probability partition does not guarantee four successive trials visit every position once. Independent selection with replacement permits repeats; without replacement is a changed process and cannot silently supply a fixed four-trial total. Link both zero positions to the probability 1/2 row.
Answer mode and expectation as different questions
The most likely single score is 0, with probability 1/2. Expectation is 1.75 although P(X=1.75)=0 because it is absent from the support. This is valid; replacing expectation with mode 0 is not. Keep separate questions for the likeliest outcome and weighted center. A particular future outcome remains undetermined until observed. Expectation lies between 0 and 5, a plausibility check that cannot alone establish correctness.
Need not be possible does not mean never possible. If Z is constantly 2 with probability 1, expectation 2 is attained every time. Separately, Y’s mode 1 still differs from expectation 4.5. Each control needs its own distribution; Z’s certainty cannot predict X’s next trial.
A sample mean need not equal expectation
A fictional four-trial record 5, 5, 0, 0 has mean 10/4=2.5 and is possible under the distribution. It is not a performed experiment. Long-run averaging under independent identically distributed repetition needs its conditions and limiting meaning, not a guarantee at four or any fixed number of trials. Keep model probabilities separate from observed frequencies; two fives out of four do not silently replace stipulated 1/4. Short samples can prompt model questions without rewriting its definition.
Recalculate a doubled-plus-one scoring rule
For Y=2X+1, scores become 1, 5, 11 with corresponding probabilities unchanged. Weighted expectation 1×1/2+5×1/4+11×1/4=18/4=4.5, independently 2×1.75+1. Add the constant to every possible score, not only the maximum. Doubling scores does not double probabilities. This is a transformation of the same variable and needs no second independent experiment. Keep X/Y tables separate to avoid a stale score row.
Y=2*X+1
possible Y: 1,5,11
E[Y]=1/2+5/4+11/4=9/2=4.5
2*E[X]+1=2*(7/4)+1=9/2
Diagnose row alignment and weights
A result 7/3 may assume three equal chances; 7 may omit weights; 4/7 may invert the fraction. Keep row labels when highlighting instead of excerpting an unidentified decimal column. Searching 1.75 may fail because of text layers or notation; return via E1 rather than treat search as probability judgment. Correct one item, then recheck probability sum and products. Without actual highlighting/search record paper checks, not automatic model validation.
Reopen and deliver three answers plus the transformation
Return to E1 and state expectation 1.75, mode 0 and fictional sample mean 2.5 before uncovering results, explaining their different objects. Recalculate Y expectation 4.5 with unchanged probabilities and retain reciprocal/equal-chance controls. Deliver distribution, partition, fictional record and arithmetic, leaving reopening pending if unperformed. New values or probabilities require recalculation; a long-run interpretation is no single-trial promise, and locating a correct number does not establish the weighting/sample distinction.
Related tutorials and sources
Combined means and weights; Binomial conditions.
Method references: OpenStax Introductory Statistics2e section4.2. Diagrams, report data, counterexamples and ledgers are original exercises.
Version checked on 17 November 2026: the Apple public lookup lists Book Isle 1.1.4, requiring iOS / iPadOS 15.6 or later; see the official App Store for published capabilities. Use self-made PDFs, EPUBs or clear text materials; bookmarks, highlights and notes depend on format and visible controls, without automatic solving or literary analysis. External preparation and paper predictions do not constitute device observations. Leave unperformed checks blank.