There is a 100-item test called the "Original Voight-Kampff Test" which has a reliability of 0.5. You have been asked to improve the reliability of this test to 0.6 by adding additional items, and the new test will be called the "Longer Voight-Kampff Test". According to the expansion factor formula, the "Longer Voight-Kampff Test" should have how many items IN TOTAL?

Question

BlackTom AI · Accepted Answer

Applying the Spearman–Brown prophecy formula for test lengthening: this formula predicts how reliability changes when you multiply the number of parallel items by a factor k. The relevant knowledge points are classical test theory (reliability increases with more parallel items) and the Spearman–Brown formula:
r_new = (k * r_old) / [1 + (k - 1) * r_old].

Start with r_old = 0.50 (original reliability) and desired r_new = 0.60. Solve for the expansion factor k.

Step-by-step algebra:
r_new = (k r_old) / [1 + (k - 1) r_old]
Multiply both sides by the denominator:
r_new [1 + (k - 1) r_old] = k r_old
Expand left side:
r_new + r_new (k - 1) r_old = k r_old
Collect terms with k on one side:
k r_new r_old - k r_old = r_new r_old - r_new
Factor k on left:
k (r_new r_old - r_old) = r_new (r_old - 1)
Recognize negatives and simplify to:
k r_old (1 - r_new) = r_new (1 - r_old)
Thus:
k = [r_new (1 - r_old)] / [r_old (1 - r_new)]

Plug in numbers:
Numerator = 0.60 * (1 - 0.50) = 0.60 * 0.50 = 0.30
Denominator = 0.50 * (1 - 0.60) = 0.50 * 0.40 = 0.20
k = 0.30 / 0.20 = 1.5

Total items required = original items * k = 100 * 1.5 = 150.

Why the correct answer is correct: a 1.5× increase in length raises reliability from 0.50 to 0.60 per the Spearman–Brown relation, so the Longer Voight-Kampff Test must have 150 items.

Why other options are incorrect:
- 100: no increase, reliability remains 0.50.
- 200: k = 2 would produce r_new = (2*0.5)/(1+0.5) = 1/1.5 ≈ 0.667, which is higher than 0.60.
- 250: k = 2.5 yields r_new ≈ 0.714, also not 0.60.

Final confirmation: the required total number of items is 150, matching the provided answer.

类似问题

With these data the specificity of the reading test to identify children with no learning difficulty is equal to:

Of the following, which is the BEST statistical practice?

Which of the following statements about Type 1 and Type 2 error is TRUE?

What effect size measure should they calculate for this dataset?

Using the same dataset, Gladly and Super Size now decide to investigate whether the average number of meatballs sold is significantly different by season (winter, spring, summer, or autumn). Assuming all assumptions are met, what statistical test would you suggest they run?

Next, Gladly looks at Cook's distance and obtains the plot below for the model m2. What does this plot indicate?

更多留学生实用工具

智能学习助手

风格化写作助手

论文查重助手

文献引用助手

课堂转译助手

课堂笔记助手

Quiz搜索助手

学校历年真题

智能刷题助手

智能匹配练习

希望你的学习变得更简单