# Trade-Offs Under Pressure - Appendix A: Initial coding schema > 前: [[@2015__MSc__Trade-Offs Under Pressure - Chapter 8 Conclusions]] | 次: [[@2015__MSc__Trade-Offs Under Pressure - Appendix B Coordinative Episodes Coding Schema]] | 全体: [[Trade-Offs Under Pressure]] ## 要約 付録 A は、IRC トランスクリプトの分析のために開発された初期コーディングスキーマ(コード名・説明・具体例の一覧)を提示する(Table 2)。このコードはパイロットインタビューで議論すべき要点を特定することに焦点を当てて作られた。さらに、パイロットインタビュー後にこのスキーマが first-order code とサブコードのグルーピングへ進化した経緯、グラフ表現のために 2 つの主要コードへ集約した経緯、そして別コーダーとの再コーディングによる信頼性(percentage agreement)の測定結果(Table 3)を記す。 ## コーディングスキーマ ### Table 2: Initial coding schema IRC トランスクリプトに対する初期コードは以下のとおり(コード名・説明・例を原文どおり転記)。 | Code | Description | Example | |---|---|---| | Asking for information, for diagnosis | A request for information, aimed at either specific individuals, or the entire channel, to aid diagnosis. | "Has something happened to that endpoint?" | | Relaying observations | The volunteering of an observation of the system's behaviour or state | "Also, the API error graph has spiked." | | Asking for clarity on a relayed observation | An attempt to repair the understanding of an observation made by someone else | "What does that line mean in the graph?" | | Giving evidence of observation | Linking to graphs, errors, code, or other artefacts to support the observation relayed | "Wait CPU per request spiked, look here: \<URL to graph>" | | Suggesting a plan of action | A suggestion of an action to take, in the form of a question or statement, with the goal of either resolving an issue, safing, or interrogating the system for more information. | "Should I shut that off?" | | Non-observation clarification | An attempt to clarify an earlier statement or repair understanding, not observation-related. | "Eng1: just to be clear, you don't want me to push this config change to staging?" | | Stating a plan of action | A declaration of what action someone is intending to do, or is already doing. | "I'm prepping a diff to turn off the sidebar and turn the blog back on." | | Asking about a plan of action | Asking others about what actions they're taking, or actions they are about to take. | "Can we hold off on doing that?" | | Asking others for feedback on plan of action | An explicit request for feedback or "sanity check" on an action someone is about to take. | "I can turn that off in isolation. How do folks feel about that?" | | Giving feedback on a plan of action | Feedback given on a plan of action, whether or not the feedback was requested. | "That looks good, go for it." | | Confirming facts or state of process | Confirming or validating an understanding of the state of the event, observation that someone has made, or other factual data. It can also be the confirmation that an action has taken place. | "Just the homepage is degraded." | | Asking about state of event | A request for the state of the event response, or asking for non-diagnostic information. | "Have we communicated with the author of the blog post?" | (Table 2. Initial coding schema — IRC トランスクリプト分析のための初期コード一覧。コード名・説明・具体例の 3 列) ### スキーマの進化 このコーディングスキーマは、パイロットインタビューを経て、first-order code とそのサブコードからなる以下のグルーピングへ進化した(原文の階層構造をそのまま転記)。 - **Diagnostic activity** - Asking for information for diagnosis - Stating a diagnostic hypothesis - Relaying observations - Giving evidence of observation - Asking for clarity on a relayed observation - **Taking action** - Suggesting a plan of action - Current or impending action - Asking other for sanity check - Asking about plan of action - Giving feedback on plan - **Confirming facts or state or process** - Asking about state of the outage response 最終的に、研究内のグラフによる表現に用いるため、first-order code はさらに 2 つの主要コードへ集約された。 - Diagnostic activity - Taking action/response ### Table 3: Intercoder Percentage Agreement コーディングシステムの信頼性を測るため、同一ドキュメントを別のコーダー(用語には詳しいが当該イベントの詳細は知らない人物)が再コーディングし、一致率(percentage agreement)を算出した。全体の一致率は 79.6% であった。 | Code | Correlates | Doesn't correlate | Total | Percent Agreement | |---|---|---|---|---| | Suggesting a plan of action | 50 | 13 | 63 | 79.3 | | Stating a diagnostic hypothesis | 58 | 34 | 92 | 63.04 | | Relaying observations | 158 | 42 | 200 | 79.0 | | Giving feedback on plan of action | 118 | 18 | 136 | 86.7 | | Giving evidence of observation | 78 | 17 | 95 | 82.1 | | Confirming facts or state of process | 159 | 41 | 200 | 79.5 | | Asking other for sanity check | 20 | 7 | 27 | 74.07 | | Asking for information for diagnosis | 100 | 21 | 121 | 82.6 | | Asking for clarity on a relayed observation | 14 | 9 | 23 | 60.8 | | Asking about state of event | 27 | 12 | 39 | 69.2 | | Asking about plan of action | 75 | 6 | 81 | 92.5 | | **合計** | **857** | **208** | **1077** | **79.6** | (Table 3. Intercoder Percentage Agreement — コードごとの一致件数・不一致件数・合計件数・一致率。全体の一致率は 79.6%) ## 関連 - [[Trade-Offs Under Pressure]] — 論文全体のハブ - [[@2015__MSc__Trade-Offs Under Pressure - Chapter 8 Conclusions]] — 前章 - [[@2015__MSc__Trade-Offs Under Pressure - Appendix B Coordinative Episodes Coding Schema]] — 次の付録(Coordinative Episodes のコーディングスキーマ) - [[@2015__MSc__Trade-Offs Under Pressure - Chapter 5 Results]] — 本コーディングスキーマが適用された IRC トランスクリプト分析の結果 ## 出典 - (Source: [[@2015__MSc__Trade-Offs Under Pressure - Appendix A Initial coding schema]], Appendix A, PDF p67-69)