{"data":{"id":198694,"slug":"postdoktor-inom-kausal-inferens-och-sprakteknologi","type":"job","title":"Postdoctoral Researcher in Causal Inference and Language Technology","description":"We have the strength of over 50,000 students and employees. Students who give hope for the future. Employees who every day contribute to Link\u00f6ping University meeting the challenges of our time. Our values are rooted in credibility, trust and security. By being bold, thinking freely and innovating, we together create\u2014through big and small actions\u2014a better future. Welcome to apply for a job with us!\n\nYour Tasks\nYou will join the research project Countering Bias in AI Methods in the Social Sciences, a collaboration between the Institute for Analytical Sociology (IAS) at Link\u00f6ping University and the Department of Computer Science and Engineering (CSE) at Chalmers University of Technology, funded by WASP-HS. The position is located at IAS. The project is at the intersection of causal inference and language technology (natural language processing, NLP).\n\nA central challenge when drawing causal conclusions from observational data is adjusting for other contextual factors (confounders). In the social sciences, these factors are often not directly measured but instead documented in unstructured text. Using text to adjust for these factors is promising, but it presents a particular risk we call treatment leakage: when the text also contains information about treatment assignment, conditioning on the text can bias the causal estimate. Your task is to collaborate with our team at CSE, who lead the method development, and then apply their methods to applied social science research questions. The methods enable researchers to use text data for causal inference while detecting and removing this source of bias.\n\nThe work combines modern language technology with causal estimation. Current directions in the project include using large language models to remove treatment-predictive information from text, evaluating and comparing estimators that correct for bias, such as Design-based Supervised Learning and Prediction-Powered Inference, and applying techniques for mechanistic interpretability, such as sparse feature circuits and the SHIFT method, so that text classifiers disregard features associated with treatment assignment.\n\nYou will help shape the research direction, lead your own studies, publish in leading scientific forums, and work closely with the project team, including serving as co-advisor to the project's doctoral student and collaborating with partners at University of Texas at Austin. Your primary focus will be on applying these methods to the effect of development projects in Africa. Secondarily on similar causal problems.\n\nAs a postdoctoral researcher, you will primarily conduct research. Teaching may also be included in your duties, but no more than one-fifth of your time.\n\nYour Qualifications\nTo be eligible for employment as a postdoctoral researcher, you must have a doctoral degree or a foreign degree assessed as equivalent to a doctoral degree. Your degree must be completed by the time the employment decision is made.\n\nIt is meritorious if your degree was completed no more than three years before the application deadline for this position. If there are special circumstances, it may be considered if your doctoral degree was completed earlier, for example due to various types of statutory leave.\n\nWe are seeking someone with a doctoral degree in social science or computational social science, or in a related quantitative field, as well as solid programming skills, for example in Python or R. You are also welcome to apply if your degree is in applied statistics, data science, machine learning or language technology with a focus on AI for Social Good or similar. Very good English skills in speaking and writing are a requirement, as the project is conducted in an international research environment. We also believe you have solid expertise in at least one of the project's two pillars\u2014causal inference or language technology\u2014as well as a genuine interest in combining the two. Experience with any of the following is highly meritorious: causal inference with observational data, text as data, large language models, mechanistic interpretability, or semiparametric and design-based estimation. Publications in leading scientific forums within social science, machine learning, language technology, statistics or computational social science are highly meritorious.\n\nYou are an independent and collaborative researcher who enjoys working across disciplinary boundaries, takes independent responsibility for a research agenda, and communicates clearly with both data scientists and social scientists.\n\nYour Workplace\nYou will work at the Institute for Analytical Sociology (IAS), which you can read more about here: https:\/\/liu.se\/organisation\/liu\/iei\/ias. At IAS you will join Professor Adel Daoud's research group, which works on artificial intelligence and causal inference for the social sciences, in close collaboration with the Data Science and AI division at Chalmers.\n\nAbout the Position\nThis is a fixed-term employment for two years with the possibility of extension for a combined employment period of no more than three years. The postdoctoral position is full-time.\n\nStart date by agreement.\n\nA security check may be conducted before an employment decision is made.\n\nSalary and Benefits\nLink\u00f6ping University applies individual and differentiated salary determination in accordance with current collective agreements (RALS\/RALS-T) and the university's guidelines for salary formation.\n\nRead more about benefits for employees: https:\/\/liu.se\/jobba-pa-liu\/formaner\n\nTrade Union Representatives\nFor information about trade union representatives, see https:\/\/liu.se\/jobba-pa-liu\/hjalp-for-sokande.\n\nApplication\nYou apply for this position by clicking the \"Apply\" button below. Your application must be received by Link\u00f6ping University no later than September 24, 2026. Applications received after the application deadline will not be considered.\n\nWe welcome applicants with different backgrounds, experiences and perspectives; this enriches and develops our activities. For us, it is self-evident to uphold the equal worth, rights and opportunities of all. Read about our work here: https:\/\/liu.se\/artikel\/lika-villkor\/.\n\nWe look forward to receiving your application!\n\nLink\u00f6ping University has procurement agreements and requests that we do not receive direct contact from staffing and recruitment companies or sellers of job advertisements.","language":"sv","is_translated":true,"title_original":"Postdoktor inom kausal inferens och spr\u00e5kteknologi","description_original":"Vi har kraften fr\u00e5n \u00f6ver 50 000 studenter och medarbetare. Studenter som ger framtidshopp. Medarbetare som varje dag bidrar till att Link\u00f6pings universitet tar sig an samtidens utmaningar. V\u00e5r v\u00e4rdegrund vilar p\u00e5 trov\u00e4rdighet, tillit och trygghet. Genom att vara modiga, t\u00e4nka fritt och g\u00f6ra nytt skapar vi tillsammans, med stora och sm\u00e5 handlingar, en b\u00e4ttre framtid. V\u00e4lkommen att s\u00f6ka jobb hos oss!\nDina arbetsuppgifter\nDu kommer att ing\u00e5 i forskningsprojektet Countering Bias in AI Methods in the Social Sciences, ett samarbete mellan Institutet f\u00f6r analytisk sociologi (IAS) vid Link\u00f6pings universitet och Institutionen f\u00f6r data- och informationsteknik (CSE) vid Chalmers tekniska h\u00f6gskola, finansierat av WASP-HS. Anst\u00e4llningen \u00e4r placerad vid IAS. Projektet ligger i sk\u00e4rningspunkten mellan kausal inferens och spr\u00e5kteknologi (natural language processing, NLP).\nEn central utmaning n\u00e4r man drar kausala slutsatser fr\u00e5n observationsdata \u00e4r att justera f\u00f6r \u00f6vriga kontextuella faktorer (confounders). Inom samh\u00e4llsvetenskapen m\u00e4ts ofga dessa faktorer inte direkt utan finns i st\u00e4llet dokumenterade i ostrukturerad text. Att anv\u00e4nda text f\u00f6r att justera f\u00f6r dessa faktorer \u00e4r lovande, men det medf\u00f6r en s\u00e4rskild risk som vi kallar behandlingsl\u00e4ckage (treatment leakage): n\u00e4r texten \u00e4ven inneh\u00e5ller information om behandlingstilldelningen kan betingning p\u00e5 texten snedvrida den kausala skattningen. Din uppgift \u00e4r att samarbeta med v\u00e5rt team vid CSE, som leder metodutvecklingen, och att d\u00e4refter till\u00e4mpa deras metoder p\u00e5 till\u00e4mpade samh\u00e4llsvetenskapliga fr\u00e5gest\u00e4llningar. Metoderna g\u00f6r det m\u00f6jligt f\u00f6r forskare att anv\u00e4nda textdata f\u00f6r kausal inferens samtidigt som denna k\u00e4lla till snedvridning uppt\u00e4cks och avl\u00e4gsnas.\nArbetet kombinerar modern spr\u00e5kteknologi med kausal skattning. Aktuella inriktningar i projektet omfattar att anv\u00e4nda stora spr\u00e5kmodeller f\u00f6r att avl\u00e4gsna behandlingsprediktiv information ur text, att utv\u00e4rdera och j\u00e4mf\u00f6ra estimatorer som korrigerar f\u00f6r snedvridning, s\u00e5som Design-based Supervised Learning och Prediction-Powered Inference, samt att till\u00e4mpa tekniker f\u00f6r mekanistisk tolkbarhet, exempelvis sparse feature circuits och SHIFT-metoden, s\u00e5 att textklassificerare bortser fr\u00e5n egenskaper som h\u00e4nger samman med behandlingstilldelningen.\nDu kommer att bidra till att forma forskningens inriktning, leda egna studier, publicera i ledande vetenskapliga fora och samarbeta n\u00e4ra med projektgruppen, bland annat genom att vara bitr\u00e4dande handledare f\u00f6r projektets doktorand och genom samarbete med partner vid University of Texas at Austin. Din prim\u00e4ra fokus kommer ligga p\u00e5 att applicera dessa metoder p\u00e5 effekten av utvecklingsprojekt i Afrika. Sekund\u00e4rt p\u00e5 liknande kausala problem.\nSom postdoktor kommer du i huvudsak bedriva forskning. \u00c4ven undervisning kan ing\u00e5 i arbetsuppgifterna, dock till h\u00f6gst en femtedel av arbetstiden.\nDina kvalifikationer\nF\u00f6r att vara beh\u00f6rig till anst\u00e4llning som postdoktor har du avlagt doktorsexamen eller har en utl\u00e4ndsk examen som bed\u00f6ms motsvara doktorsexamen. Din examen ska vara avlagd senast vid tidpunkten d\u00e5 anst\u00e4llningsbeslutet fattas.\nDet \u00e4r meriterande om din examen \u00e4r avlagd h\u00f6gst tre \u00e5r f\u00f6re sista ans\u00f6kningsdag f\u00f6r denna anst\u00e4llning. Om det finns s\u00e4rskilda sk\u00e4l kan den komma i fr\u00e5ga som avlagt doktorsexamen tidigare, exempelvis vid olika typer av lagstadgade ledigheter.\nVi s\u00f6ker dig som har doktorsexamen inom samh\u00e4llsvetenskap eller ber\u00e4kningssamh\u00e4llsvetenskap (computational social science), eller inom ett n\u00e4rliggande kvantitativt \u00e4mne, samt gedigna programmeringskunskaper, exempelvis i Python eller R. Du \u00e4r ocks\u00e5 v\u00e4lkommen att s\u00f6ka om din examen \u00e4r inom till\u00e4mpade delar av statistik, datavetenskap, maskininl\u00e4rning eller spr\u00e5kteknologi och har en inriktning mot AI for Social Good eller liknande. Mycket goda kunskaper i engelska i tal och skrift \u00e4r ett krav, eftersom projektet bedrivs i en internationell forskningsmilj\u00f6. Vi tror ocks\u00e5 att du har gedigen utbildning inom minst en av projektets tv\u00e5 pelare, kausal inferens eller spr\u00e5kteknologi, samt ett genuint intresse f\u00f6r att f\u00f6rena de tv\u00e5. Erfarenhet av n\u00e5got av f\u00f6ljande \u00e4r starkt meriterande: kausal inferens med observationsdata, text som data, stora spr\u00e5kmodeller, mekanistisk tolkbarhet eller semiparametrisk och designbaserad skattning. Publikationer i ledande vetenskapliga fora inom samh\u00e4llsvetenskap, maskininl\u00e4rning, spr\u00e5kteknologi, statistik eller ber\u00e4kningssamh\u00e4llsvetenskap \u00e4r mycket meriterande.\nDu \u00e4r en sj\u00e4lvg\u00e5ende och samarbetsinriktad forskare som trivs med att arbeta \u00f6ver disciplingr\u00e4nser, tar sj\u00e4lvst\u00e4ndigt ansvar f\u00f6r en forskningsagenda och kommunicerar tydligt med s\u00e5v\u00e4l datavetare som samh\u00e4llsvetare.\nDin arbetsplats\nDu kommer att arbeta p\u00e5 Institutet f\u00f6r analytisk sociologi (IAS), som du kan l\u00e4sa mer om h\u00e4r: https:\/\/liu.se\/organisation\/liu\/iei\/ias. Vid IAS ing\u00e5r du i professor Adel Daouds forskargrupp, som arbetar med artificiell intelligens och kausal inferens f\u00f6r samh\u00e4llsvetenskapen, i n\u00e4ra samarbete med avdelningen f\u00f6r Data Science och AI vid Chalmers.\nOm anst\u00e4llningen\nAnst\u00e4llningen \u00e4r en tidsbegr\u00e4nsad anst\u00e4llning p\u00e5 tv\u00e5 \u00e5r med m\u00f6jlighet till f\u00f6rl\u00e4ngning med en sammanlagd anst\u00e4llningstid p\u00e5 h\u00f6gst tre \u00e5r. Anst\u00e4llningen som postdoktor \u00e4r p\u00e5 heltid.\nTilltr\u00e4de enligt \u00f6verenskommelse.\nS\u00e4kerhetskontroll kan komma att genomf\u00f6ras innan beslut om anst\u00e4llning fattas.\nL\u00f6n och f\u00f6rm\u00e5ner\nLink\u00f6pings universitet till\u00e4mpar individuell och differentierad l\u00f6nes\u00e4ttning och sker i enlighet med g\u00e4llande kollektivavtal (RALS\/RALS-T) samt universitetets riktlinjer f\u00f6r l\u00f6nebildning.\nL\u00e4s mer om f\u00f6rm\u00e5ner f\u00f6r anst\u00e4llda https:\/\/liu.se\/jobba-pa-liu\/formaner\nFackliga kontaktpersoner\nInformation om fackliga kontaktpersoner, se https:\/\/liu.se\/jobba-pa-liu\/hjalp-for-sokande.\nAns\u00f6kan\nDu s\u00f6ker denna anst\u00e4llning genom att klicka p\u00e5 knappen \"Ans\u00f6k\" nedan. Din ans\u00f6kan ska vara Link\u00f6pings universitet tillhanda senast den 24 september 2026. Ans\u00f6kan som inkommer efter sista ans\u00f6kningsdag beaktas ej.\nVi v\u00e4lkomnar s\u00f6kande med olika bakgrund, erfarenheter och perspektiv, det berikar och utvecklar v\u00e5r verksamhet. F\u00f6r oss \u00e4r det sj\u00e4lvklart att v\u00e4rna om allas lika v\u00e4rde, r\u00e4ttigheter och m\u00f6jligheter. L\u00e4s om v\u00e5rt arbete med https:\/\/liu.se\/artikel\/lika-villkor\/.\nV\u00e4lkommen med din ans\u00f6kan!\nLink\u00f6pings universitet har upphandlade avtal och undanber oss direktkontakt fr\u00e5n bemannings- och rekryteringsf\u00f6retag samt f\u00f6rs\u00e4ljare av platsannonser.","price":null,"currency":"SEK","status":"active","noindex":true,"location":{"address":null,"full_address":null,"city":"Norrk\u00f6ping","country":"SE","latitude":58.590692121759,"longitude":16.180383818356},"metadata":{"region":"\u00d6sterg\u00f6tlands l\u00e4n","duration":"6 m\u00e5nader eller l\u00e4ngre","employer":"Link\u00f6pings Universitet","postcode":"60221","positions":1,"profession":"Universitets- och h\u00f6gskolelektor","salary_type":"Fast m\u00e5nads- vecko- eller timl\u00f6n","employer_url":"https:\/\/liu.se\/","scope_of_work":"100\u2013100 %","working_hours":"Heltid","contact_person":"Adel Daoud","employment_type":"full_time","occupation_field":"Pedagogik","employer_workplace":"Link\u00f6pings Universitet","experience_required":true,"employment_type_label":"Vanlig anst\u00e4llning"},"user_id":null,"is_sponsored":false,"views_count":0,"ai_views_count":6,"bot_views_count":2,"ai_vendor_counts":{"other":2,"openai":1,"anthropic":3},"visibility":"public","submission_source":null,"submission_ai_name":null,"has_owner_email":true,"can_contact_owner":true,"phone":null,"owner_email":"adel.daoud@liu.se","images":[],"published_at":"2026-09-07T15:27:34+00:00","expires_at":"2026-09-24T23:59:59+00:00","created_at":"2026-09-08T03:54:20+00:00","updated_at":"2026-09-16T03:00:10+00:00"}}