bias

5
Bias Bias yang dimaksud di halaman ini ialah keberpihakan dari peneliti atau penerbit laporan penelitian ke kesimpulan dan rekomendasi yang tidak valid, baik yang tidak disengaja ataupun yang tidak diakui. Bias peneliti dapat terjadi pada semua tahap penelitian (persiapan, pengumpulan data, pengolahan data, penafsiran data dan pelaporan) dan dapat dideteksi di laporan penelitian yang dipublikasi melalui pertemuan ilmiah atau berkala ilmiah. Bias penerbit dari penyelenggara pertemuan ilmiah atau penerbit berkala ilmiah dapat mempengaruhi ke arah yang keliru pembuatan keputusan pendana proyek penelitian, pembuatan kebijakan publik dan perilaku masyarakat. Kecenderungan yang keliru ini sering diperparah oleh “bias media” – kecenderungan penerbit media komunikasi masa untuk lebih banyak atau lebih cepat mempublikasikan kesimpulan dan rekomendasi penelitian yang mendukung hipotesis walaupun tidak valid. Bias Penafsiran Data The placebo effect in medicine, where getting an inert (e.g. sugar) pill has a large positive effect. Many believe that there are often large positive effects apparently simply from the expectation created in the patient: if true, this is the placebo effect, where the intervention in fact has no material effect, but the belief by the participant does.

Upload: aji-wibowo

Post on 15-Nov-2015

12 views

Category:

Documents


2 download

DESCRIPTION

bias dalam melakukan penelitian

TRANSCRIPT

BiasBias yang dimaksud di halaman ini ialah keberpihakan dari peneliti atau penerbit laporan penelitian ke kesimpulan dan rekomendasi yang tidak valid, baik yang tidak disengaja ataupun yang tidak diakui. Bias peneliti dapat terjadi pada semua tahap penelitian (persiapan, pengumpulan data, pengolahan data, penafsiran data dan pelaporan) dan dapat dideteksi di laporan penelitian yang dipublikasi melalui pertemuan ilmiah atau berkala ilmiah. Bias penerbit dari penyelenggara pertemuan ilmiah atau penerbit berkala ilmiah dapat mempengaruhi ke arah yang keliru pembuatan keputusan pendana proyek penelitian, pembuatan kebijakan publik dan perilaku masyarakat. Kecenderungan yang keliru ini sering diperparah oleh bias media kecenderungan penerbit media komunikasi masa untuk lebih banyak atau lebih cepat mempublikasikan kesimpulan dan rekomendasi penelitian yang mendukung hipotesis walaupun tidak valid.Bias PenafsiranData The placebo effect in medicine, where getting an inert (e.g. sugar) pill has a large positive effect. Many believe that there are often large positive effects apparently simply from the expectation created in the patient: if true, this is the placebo effect, where the intervention in fact has no material effect, but the belief by the participant does. Although often transmitted from the doctors expectancies, it may be independent of the doctor. It may show particularly strongly in side-effects, where the number and severity of side-effects may be three times larger when patients are warned about the possibility in both groups that get the active treatment and in the placebo group. However as noted above, some do not believe any such effect exists. The Hawthorne effect: simply of being studied. Aspects of this suggest that the effect did not depend on the particular expectation of the researchers, but that being studied caused the improved performance. This might be because attention made the workers feel better; or because it caused them to reflect on their work and reflection caused performance improvements, or because the experimental situation provided them with performance feedback they didnt otherwise have and this extra information allowed improvements. The John Henry effect (Zdep & Irvine; 1970) is the opposite of the Hawthorne effect: it is when a supposedly control group, that gets no intervention, compares themselves to the experimental group and through extra effort gets the same effects or results. A kind of counter-suggestibility. http://w2.xrefer.com/entry/151552 The halo effect of uncontrolled novelty: the participant performs differently at first because of the novelty of the treatment which may change their expectation, or simply cause them to be more alert or otherwise perform differently. The experimenter is not important, but a materially unjustified belief, perhaps from other social media, may be (e.g. participants think the technology / educational intervention is wonderful and that belief is the real cause of raised outcomes); or else simply the novelty rather than belief matters, if it operates through (say) attention rather than through expectancies. Experimenter effects. Specific expectations acquired, consciously or not, from the researcher. Some experimenter effects have been demonstrated equally in positive and negative directions. Rosenthal (1966) describes experimentally tested experimenter effects in behavioral research, which is summarised by Rosenthal & Jacobson (1992). Prophesying a difference caused research assistants to create an effect, and this could be done equally in either direction (i.e. can create a positive or negative effect this way). This was done where the experimental task being manipulated required judgements by the nominal participants. However this was about one tenth the size of the effect prophesied, so it would be quite wrong to describe this as seeing what you expect: it would be more accurate to suggest that experimenters could influence subjects on marginal cases and so systematically bias (only) within the range of experimental noise. However if stooges acting as the first subjects behaved differently, this overrode and created a more effective expectation (and consequent effect on real subjects). Such effects have also been demonstrated in animal experiments and on learning and IQ tests/tasks at least sometimes. Jastrows effect on factory work was much bigger: here an explicit expectation about performance was transmitted and turned out to change output by a factor of three. The Pygmalion effect or expectancy advantage is that of a self-fulfilling prophecy. Teachers expectations of pupils can strongly affect (by about a factor of two over a year) the amount of development they show.Source (25 Sept 2012): http://www.psy.gla.ac.uk/~steve/hawth.htmlDefinitions of interpretation biasesSource (12 Feb 2013): http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1126323/Confirmation biasevaluating evidence that supports ones preconceptions differently from evidence that challenges these convictionsRescue biasdiscounting data by finding selective faults in the experimentAuxiliary hypothesis biasintroducing ad hoc modifications to imply that an unanticipated finding would have been otherwise had the experimental conditions been differentMechanism biasbeing less sceptical when underlying science furnishes credibility for the dataTime will tell biasthe phenomenon that different scientists need different amounts of confirmatory evidenceOrientation biasthe possibility that the hypothesis itself introduces prejudices and errors and becomes a determinate of experimental outcomesBiasPenerbitProteus Phenomenon: menerbitkan lebih awal laporan penelitian yang kesimpulannya kontroversial. (10 Jan 12: http://en.wikipedia.org/wiki/Proteus#cite_note-3)Positive-outcome Bias: menerbitkan lebih sering laporan penelitian yang kesimpulannya mendukung hipotesis. (10 Jan 12: http://www.skepdic.com/posoutbias.html)