44:["$","title",null,{"children":"서명 먼저 하기 데이터 조작 흔적 | taekie | Observable"}] 4c:[{"id":"b02025169b718158a7f1584c4edfc225a096f5dd3bb020854f2bd9891bab4d617753e2f0dfefccba8642994a9a21d18755e577fa89a3ce9f4dccdf3ba9262b03","href":"https://static.observableusercontent.com/files/b02025169b718158a7f1584c4edfc225a096f5dd3bb020854f2bd9891bab4d617753e2f0dfefccba8642994a9a21d18755e577fa89a3ce9f4dccdf3ba9262b03","name":"Data-Table 1.csv","type":"text/csv","lastModified":1629795894817.025,"size":961974},{"id":"693a3ae6b42d2cec8997013f3a86b04c5e3752934c5b9e1f1fd9e2767449876b70d3251f293bd140b66475e41f8a639155ffca091265ad81dfc389491481d2ae","href":"https://static.observableusercontent.com/files/693a3ae6b42d2cec8997013f3a86b04c5e3752934c5b9e1f1fd9e2767449876b70d3251f293bd140b66475e41f8a639155ffca091265ad81dfc389491481d2ae","name":"IMG_9598.jpg","type":"image/jpeg","lastModified":1629934925588.821,"size":1821338}] 48:{"id":"754bf5fa2184db07","slug":"signing_at_the_beginning","title":"서명 먼저 하기 데이터 조작 흔적","thumbnail":"1ba53a2d0629fa825ef2f7d22b8117e698027d661c4e83e0fa8a21a20dba0e4e","access_level":"public","create_time":"$D2021-08-24T09:04:22.009Z","update_time":"$D2021-08-26T00:03:47.233Z","view_time":null,"first_public_time":"$D2021-08-25T23:35:20.023Z","has_unpublished_edits":null,"fork":false,"stars":0,"starred":false,"owner":{"type":"team","id":"b7e5b2f89b823a97","login":"taekie","name":"taekie","avatar_url":"https://avatars-next.observableusercontent.com/be34fc83e1c21b2cc0145ceebbde48e5fcfa280c49aa41dbd746a1c3c1847b81","tier":"pro_2024","bio":"","followed_by_you":false,"follows_you":false},"version":531,"publish_version":531,"role":null,"authored":false,"authors":[{"type":"individual","id":"8bc558cbad6934ee","login":"taekie","name":"taekie","avatar_url":"https://avatars-next.observableusercontent.com/be34fc83e1c21b2cc0145ceebbde48e5fcfa280c49aa41dbd746a1c3c1847b81","tier":null,"bio":"","followed_by_you":false,"follows_you":false}],"members":[{"type":"individual","id":"8bc558cbad6934ee","login":"taekie","name":"taekie","avatar_url":"https://avatars-next.observableusercontent.com/be34fc83e1c21b2cc0145ceebbde48e5fcfa280c49aa41dbd746a1c3c1847b81","tier":null,"role":"editor"}],"comments":[],"custom_thumbnail":false,"trashed":false,"lapsed":null,"subscribed":null,"forks":0,"collection":null,"type":"notebook","body":{"titling":"auto","title":"서명 먼저 하기 데이터 조작 흔적","cells":[{"id":0,"value":"md`# 서명 먼저 하기 데이터 조작 흔적`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":323,"value":"md`### 서명 먼저 하기\n댄 애리얼리의 <[거짓말하는 착한 사람들](https://book.naver.com/bookdb/book_detail.nhn?bid=6968107)> 책에 인용된 [논문](https://www.pnas.org/content/109/38/15197)의 실험\\n\n

\n보통 서류의 맨 뒤에 \"내가 제공하는 정보가 사실임을 맹세한다\"라는 문구와 서명이 있는데 그것 맨 위로 옮기면 거짓이 줄어 든다는 가설과 그 실험이다.
\n자동차 보험에 운행 거리가 적으면 보험 할인을 더 받을 수 있어 조금씩 속이는 경향이 있는데 서명 먼저 하기가 효과가 있는지 확인하는 현장 실험을 하였다.\\n\ndatacolada에서 원 데이터를 분석하고 조작 흔적을 찾아 공개하였다. https://datacolada.org/98`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":320,"value":"md`#### 데이터`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":144,"value":"md`데이터 출처 https://osf.io/3javq/ 에서 excel 파일을 csv로 변환`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":4,"value":"raw = d3.csvParse(await FileAttachment(\"Data-Table 1.csv\").text(), d3.autoType)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":317,"value":"Inputs.table(raw)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":352,"value":"md`\n* OMR Version - 서명을 먼저했는지 뒤에 했는 지 구분\n* Odom 1, Odom 2,... - 계약자의 차량 수로 대부분 1대지만 최대 4대까지 있다.\n* Odom Reading - 차량의 주행 기록계\n * (previous) - 초기 마일리지\n * (update) - 마지막 마일리지\n * diff - 둘의 차이가 운행 거리이고 이 거리가 적으면 할인이 되어 거짓으로 보고하는 유인이 된다.\\n\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":462,"value":"md`#### 첫번째 차량 주행 거리 (N=${data.length})\\n\n첫번째 차량의 초기 마일리지와 마지막 마일리지, 두 차이인 운행 거리만 추출`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":18,"value":"data = raw.map(d => ({\n sign: d[\"OMR Version\"],\n odom1: d[\"Odom Reading 1 (Previous)\"],\n odom1update: d[\"Odom Reading 1 (Update)\"],\n diff1: d[\"Odom Reading 1 (Update)\"] - d[\"Odom Reading 1 (Previous)\"]\n}))","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":225,"value":"md`#### 서명 위치에 따른 평균 차량 운행 거리\n서명을 위에 했을 때 아래했을 때 보다 신고한 운행 거리가 11%정도 차이가 있다
\n논문과 책에서 서명 먼저 하기가 거짓을 줄이는 데 효과가 있다는 근거.
\n하지만 이후 다른 연구에서 실험을 재현 못해 데이터 요구하게 됨
`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":176,"value":"embed({\n $schema: 'https://vega.github.io/schema/vega-lite/v5.json',\n width: 600,\n height: 100,\n\n data: {\n values: data,\n name: \"source\"\n },\n mark: \"bar\",\n encoding: {\n color: {\n field: \"sign\",\n type: \"nominal\"\n },\n y: {\n field: \"sign\",\n type: \"nominal\"\n },\n x: {\n aggregate: \"mean\",\n field: \"diff1\",\n type: \"quantitative\",\n scale: { domain: [0, 30000] }\n }\n },\n layer: [\n {\n mark: { type: \"bar\", tooltip: { content: \"data\" } }\n },\n {\n mark: {\n type: \"text\",\n tooltip: { content: \"data\" },\n align: \"right\",\n baseline: \"middle\",\n dx: -5,\n dy: 0,\n fill: \"#fff\"\n },\n encoding: {\n text: { aggregate: \"mean\", field: \"diff1\", format: \",.0f\" }\n }\n }\n ],\n config: {\n line: {\n point: true\n },\n scale: {\n useUnaggregatedDomain: true\n }\n }\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":228,"value":"md`#### 초기 마일리지 분포\n5000마일 미만 신규 차량이 많지만 그 외는 정규 분포에 가깝다 `","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":293,"value":"embed({\n $schema: 'https://vega.github.io/schema/vega-lite/v5.json',\n width: 600,\n height: 100,\n\n data: {\n values: data,\n name: \"source\"\n },\n // transform: [{ sample: 1000 }],\n mark: \"tick\",\n encoding: {\n color: {\n field: \"sign\",\n type: \"nominal\"\n },\n opacity: { value: 0.05 },\n y: {\n field: \"sign\",\n type: \"nominal\"\n },\n x: {\n field: \"odom1\",\n type: \"quantitative\"\n }\n },\n config: {\n line: {\n point: true\n },\n scale: {\n useUnaggregatedDomain: true\n }\n }\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":21,"value":"viewof view222 = embed({\n $schema: 'https://vega.github.io/schema/vega-lite/v5.json',\n width: 500,\n height: 300,\n\n data: {\n values: data,\n name: \"source\"\n },\n mark: { type: \"bar\", tooltip: { content: \"data\" } },\n\n encoding: {\n x: {\n bin: { maxbins: 100, extent: [0, 300000] },\n field: \"odom1\",\n type: \"quantitative\"\n },\n y: {\n aggregate: \"count\"\n }\n },\n config: {\n line: {\n point: true\n },\n scale: {\n useUnaggregatedDomain: true\n }\n }\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":239,"value":"md`### 1. 차량 운행 거리 분포가 인위적임\n마지막 마일리지에서 초기 마릴리지를 뺀 차량 운행 거리가 0부터 50,000에서 딱 끊김. 50,000 이상은 보험 가입을 거부했나?
사람들이 차를 이용하는 행태가 비슷비슷하게 평균에 몰려있고 극단적으로 많이 운행한 사람도 조금씩 있는 정규 분포일 것 같은데 **아래 균등 랜덤 생성한 가짜 데이터 분포에 가깝다**.
\n초기 운행 기록에 0-50,000 사이 랜덤 값을 더해서 두번째 마일리지를 조작했을 것으로 추측`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":187,"value":"embed({\n $schema: 'https://vega.github.io/schema/vega-lite/v5.json',\n width: 600,\n height: 100,\n\n data: {\n values: data,\n name: \"source\"\n },\n // transform: [{ sample: 1000 }],\n mark: \"tick\",\n encoding: {\n color: {\n field: \"sign\",\n type: \"nominal\"\n },\n opacity: { value: 0.1 },\n y: {\n field: \"sign\",\n type: \"nominal\"\n },\n x: {\n field: \"diff1\",\n type: \"quantitative\"\n }\n },\n config: {\n line: {\n point: true\n },\n scale: {\n useUnaggregatedDomain: true\n }\n }\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":108,"value":"embed({\n $schema: 'https://vega.github.io/schema/vega-lite/v5.json',\n width: 500,\n height: 300,\n\n data: {\n values: data,\n name: \"source\"\n },\n mark: \"bar\",\n encoding: {\n x: {\n bin: { maxbins: 50 },\n field: \"diff1\",\n type: \"quantitative\"\n },\n y: {\n aggregate: \"count\"\n }\n },\n config: {\n line: {\n point: true\n },\n scale: {\n useUnaggregatedDomain: true\n }\n }\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":251,"value":"md`#### 균등 분포 랜덤`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":246,"value":"uniformRandom = data.map(d => ({ diff1: d3.randomInt(0, 50000)() }))","pinned":true,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":243,"value":"embed({\n $schema: 'https://vega.github.io/schema/vega-lite/v5.json',\n width: 500,\n height: 300,\n\n data: {\n values: uniformRandom,\n name: \"source\"\n },\n mark: \"bar\",\n encoding: {\n x: {\n bin: { maxbins: 50 },\n field: \"diff1\",\n type: \"quantitative\"\n },\n y: {\n aggregate: \"count\"\n }\n },\n config: {\n line: {\n point: true\n },\n scale: {\n useUnaggregatedDomain: true\n }\n }\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":262,"value":"md`### 2. 마지막 마일리지엔 반올림이 보이지 않음\n마지막 3자리 000-999 빈도. 반올림(000,100,200...)수가 눈에 띄지 않음. 사람이 적지 않고 랜덤 생성된 수라고 생각할 수 있다.`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":117,"value":"lastDigit1u = data.map(d => ({ digit: d.odom1update % 1000 }))","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":127,"value":"embed({\n $schema: 'https://vega.github.io/schema/vega-lite/v5.json',\n width: 500,\n height: 300,\n\n data: {\n values: lastDigit1u,\n name: \"source\"\n },\n mark: { type: \"point\", tooltip: { content: \"data\" } },\n encoding: {\n fill: {\n value: \"#bbb\",\n condition: {\n test: \"datum['bin_maxbins_1000_digit']%100===0\",\n value: \"#2883F7\"\n }\n },\n color: {\n value: \"transparent\"\n },\n x: {\n bin: { maxbins: 1000 },\n field: \"digit\",\n type: \"quantitative\",\n axis: {\n values: [0, 100, 200, 300, 400, 500, 600, 700, 800, 900]\n }\n },\n y: {\n aggregate: \"count\"\n }\n },\n config: {\n line: {\n point: true\n },\n scale: {\n useUnaggregatedDomain: true\n }\n }\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":259,"value":"md`#### 초기 마일리지 마지막 3자리\n6자리를 다 적기는 귀찮으니 반올림하여 000이 가장 많고 100,200.. 이 눈에 띄게 많음`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":115,"value":"lastDigit1 = data.map(d => ({ digit: d.odom1 % 1000 }))","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":119,"value":"viewof view11 = embed({\n $schema: 'https://vega.github.io/schema/vega-lite/v5.json',\n width: 500,\n height: 300,\n\n data: {\n values: lastDigit1,\n name: \"source\"\n },\n transform: [\n {\n bin: true,\n field: \"digit\",\n as: \"binnedDigit\"\n }\n ],\n\n mark: { type: \"point\", tooltip: { content: \"data\" } },\n encoding: {\n fill: {\n value: \"#bbb\",\n condition: {\n test: \"datum['bin_maxbins_1000_digit']%100===0\",\n value: \"#2883F7\"\n }\n },\n color: {\n value: \"transparent\"\n },\n x: {\n bin: { maxbins: 1000 },\n field: \"digit\",\n type: \"quantitative\",\n axis: {\n values: [0, 100, 200, 300, 400, 500, 600, 700, 800, 900]\n }\n },\n y: {\n aggregate: \"count\"\n }\n },\n config: {\n line: {\n point: true\n },\n scale: {\n useUnaggregatedDomain: true\n }\n }\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":22,"value":"embed = require(\"vega-embed@6\")","pinned":true,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","tool":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"}],"files":"$W4c","stdlib":"1"}} 24:null