問8
頻出度★★★
2種類のデータの関係性を表すことを目的として用いるものはどれか。
平成24年度 秋期 ITパスポート試験 問8
ストラテジ系企業活動 › 業務分析・データ利活用
This question's explanation and your answers so far.
解説
問 8
Incorrectア
A decision table lists the possible combinations of conditions vertically and organizes, in table form, the action (processing) to be executed for each combination using Y/N or X marks. It is a tool for checking complex conditional branches for gaps and duplicates, and for listing specifications and test cases exhaustively; it does not show the relationship between two kinds of numerical data.
- イ
正解 The correct answer is イ, the scatter plot. In a scatter plot, one kind of data (for example, temperature) is taken on the horizontal axis and the other kind of data (for example, the number of products sold) on the vertical axis, and each pair of values obtained from the same object is plotted as one point on the plane. By looking at how the plotted points are scattered (clustering upward to the right = positive correlation, downward to the right = negative correlation, spread out at random = no correlation), you can visually judge whether there is a relationship between the two kinds of data. It is exactly a chart whose purpose is “to show the relationship between two kinds of data”, and as one of the seven basic QC tools it is used in quality control and statistical data analysis.
Incorrectウ
A cause-and-effect diagram (fishbone diagram, Ishikawa diagram) places one characteristic (the result) at the tip of the arrow that forms the backbone, and writes out the contributing factors toward it hierarchically as large bones, medium bones, and small bones. It is meant for classifying and organizing causes from viewpoints such as people, machines, materials, and methods; it is not a chart that pairs two kinds of data to look at their relationship.
Incorrectエ
A Pareto chart tallies items such as defects or complaints by category, arranges them as bars in descending order of count or amount, and overlays a line showing the cumulative percentage. It is used to identify the small number of key items that account for most of the total (ABC analysis, priority management); what it handles is one kind of data per item, so it does not show the relationship between two kinds of data.
ここがポイント
- Memorize how the seven basic QC tools are used differently, by their purpose: scatter plot = see the correlation between two kinds of data, Pareto chart = narrow down the key items, cause-and-effect diagram = list up the contributing factors, histogram = see the distribution of variation, control chart = see the stability of a process.
- What can be read from a scatter plot is only correlation, not causation. Even if the points line up going upward to the right, you cannot conclude that one of them is the cause of the other; be careful about this point.
用語