{"args":{},"checks":[{"abs_tol":null,"bind":"any","column":null,"failure_class":null,"field":"value","interval":false,"key":null,"kind":"answer_backed","min_lre":null,"note":"every number of the answer is stated by a tool call","open_interval":false,"rel_tol":null,"required":true,"source_id":null,"table":null,"tool":null,"value":null,"weight":1.0,"where":null},{"abs_tol":null,"bind":"answer_call","column":null,"failure_class":null,"field":"value","interval":true,"key":null,"kind":"answer_figure_close","min_lre":null,"note":"QRData eval.py: strictly inside gold \u00b1 3% (a '%' gold divided by 100), on the answer figure after meta.answer_format","open_interval":true,"rel_tol":null,"required":true,"source_id":"qrdata.gold.363","table":null,"tool":null,"value":[3.1913,3.3887],"weight":1.0,"where":null}],"datasets":{"app_engagement_push":{"columns":null,"description":"A study is conducted to measure the effect of a marketing push on user engagement, specifically in-app purchases. Some customers who were assigned to receive the push are not receiving it, because they probably have an older phone that doesn\u2019t support the kind of push the marketing team designed.\nThe dataset app_engagement_push.csv contains records for 10,000 random customers. Each record includes whether an in-app purchase was made (in_app_purchase), if a marketing push was assigned to the user (push_assigned), and if the marketing push was successfully delivered (push_delivered).","file_sha256":"3c3a97cca6fdd96f6856f7856b5f8bc786a31d1a9be78c9f610ef3070783e970","format":"csv","frame_sha256":null,"generator":null,"kind":"file","licence":"CC BY-NC 4.0","member":"data/app_engagement_push.csv","object":null,"params":{},"path":null,"read_kw":{},"records":null,"role":"agent","sha256":"3c3a97cca6fdd96f6856f7856b5f8bc786a31d1a9be78c9f610ef3070783e970","transforms":[],"url":"https://raw.githubusercontent.com/xxxiaol/QRData/de450af45ff7101b328bb064c6b475f73414a7ed/benchmark/data.zip"}},"difficulty":"M","id":"qrdata_socsci.363","level":"T2","meta":{"agent_reach":{"blocked_by":[],"dtypes":{"app_engagement_push":{"in_app_purchase":"int64","push_assigned":"int64","push_delivered":"int64"}},"named":{"outcome":["in_app_purchase"],"treatment":["push_delivered"]},"reachable":true,"rule":"magi_bench.reach/1"},"answer_format":{"figure_transform":"identity","round_to":2,"unit_scale":1.0},"answer_format_from":{"figure_transform":null,"round_to":"two decimal places","unit_scale":"1.0: an estimator's figure is on the gold's own scale"},"cluster":"app_engagement_push.csv","expected_tool":"infer.iv_2sls","family":"numerical","headline_keys":{"infer.iv_2sls":["coef.push_delivered"]},"naive":{"dataset":"app_engagement_push","treatment":"push_delivered","y":"in_app_purchase"},"native":{"answer":"3.29","keywords":["Causality","Observational data","Regression analysis","Local average treatment effect","Instrumental variables"],"multiple_choices":null,"question_type":"numerical","reference":"Causal Inference for the Brave and True 9"},"native_id":363,"reference":{"cite":1,"note":"Caminho de refer\u00eancia inferido \u00e0 m\u00e3o.","steps":[{"args":{"dataset":"app_engagement_push","endog":["push_delivered"],"instruments":["push_assigned"],"y":"in_app_purchase"},"tool":"infer.iv_2sls","why":"2SLS of in-app purchases on delivery instrumented by assignment: the LATE."}]},"reference_path":{"calls":[{"args":{"dataset":"app_engagement_push","endog":["push_delivered"],"instruments":["push_assigned"],"y":"in_app_purchase"},"tool":"infer.iv_2sls"}],"needs":[],"provenance":"hand_inferred","quotes":{},"reference_datasets":[],"source_id":null},"reference_why":"2SLS of in-app purchases on delivery instrumented by assignment: the LATE."},"question":"What is the Local Average Treatment Effect (LATE) of receiving the marketing push on in-app purchases, as estimated using linear regression and instrumental variable, rounded to two decimal places?","requires":[],"runner":null,"sources":["qrdata.gold.363","qrdata.eval_py"],"split":"public","suite":"qrdata_socsci","tags":["econ-bench","external","qrdata","brave_and_true","numerical","reference_path","agent_reachable","naive"],"target":"agent","tool":null,"truth":"QRData gold: benchmark/QRData.json item 363 at de450af45ff7101b328bb064c6b475f73414a7ed"}