Biases, Ascertainment biases, Innovation and Project Design.
Biases:
Biases are systematic (or let’s call them inherent) errors or limitations in the way an idea is conceived, a comment is responded, a person is acknowledged, a population is treated and sometimes even how governments deal with priorities. In general, these biases reflect a person’s (or a group’s) ability to remain agnostic and think independently.
In technical field, it is every evident how a research project is outlined, or the information is collected, processed, analyzed, or interpreted. The biases may be rooted to personal beliefs, work environment, level of independence, influences, patterns, previous experiences, or the way someone envision himself/herself or his/her standing in the future.
Biases usually arise when individuals or groups only look for or consider certain types of ideas, data, or perspectives, while ignoring or dismissing others. Regardless, biases can kill creativity by limiting the scope of what is considered possible or valuable in a particular field.
Ascertainment Biases:
Ascertainment biases can simply be explained as biased in sampling or data/information collections.
Ascertainment biases can limit the diversity of voices and perspectives that are included in creative fields, leading to a lack of innovation and originality. Ascertainment biases occur when researchers and funding agencies/investors only want to look for or consider certain types of information/data, while ignoring or excluding other evidence/information/data that may be relevant or important. Sometimes people just don’t want to explore in new domains and keep working on the ‘known’ tracks. That is the hidden ascertainment bias.
Remember when you applied for a grant for ‘new’ project or discussed a ‘new’ idea with your supervisor. Not a rule, by there are trends that people with significant professional experiences have significant professional biases; sometimes because they are too certain and opinionated, and sometimes because they never tried (or don’t want) to explore anything else (anymore). Fun part: In academic research, sometimes it is unaffordable to remain unbiased, the funding doesn’t allow you to go big and go fair. Plus, you got to please study sections. Anyways, these biases can arise in any field of study or research, and they can have a significant impact on the validity and reliability of findings.
Biases in technical project designs:
Ascertainment biases can have a significant impact on the fields of science and technology and may lead to a lack of diversity and innovation in research and development.
When a clinical study only recruits participants (or requires samples) from a certain age, gender, or demographic group, the results may not be applicable to the broader population. Similar is applicable for a basic researcher when he/she doesn’t include references from different perspective and ‘unintentionally’ keeps the study tilted towards certain outcomes- that often fits a specific narrative. The available literature is full of such studies and reports which are useful but doesn’t provide the overall picture of the subject matter. For example, if a group of scientists only consider research that supports a particular theory, they may miss out on potentially groundbreaking discoveries that challenge their assumptions. Since basic research is ‘slow’ and routinely reviewed and summarized, these biases (unless they are misconducts) are neutralized over the time.
Similarly, in technology, ascertainment biases can lead to the development of products or systems that do not fully meet the needs of all users. For example, if a particular technology is designed only with certain users in mind, it may not be accessible or useful for others (unless it is intended not to be accessed by all the groups, let’s say- Tinder!). This can limit the potential impact of new technologies and may result in missed opportunities for innovation and growth. For example, if a common use technology such as WhatsApp or Facebook end up being inaccessible to certain groups, reducing the utility and hence reducing the monetary translation of the efforts put to build the technology.
Ascertainment biases can also arise in non-technical areas such as if there is a bias in the way information is collected or analyzed for generic studies. For example, if an online survey is conducted in a way that is more likely to elicit certain types of responses, the results may not accurately represent the views of the entire population. Similarly, if an artist only produces work that is like what has been successful in the past, they may miss out on exploring new techniques or styles.
Remedy:
To overcome ascertainment biases and promote creativity, it is important to actively seek out and consider diverse perspectives and to remain open to new and unconventional ideas. By embracing diversity and being willing to take risks and explore new territory, creative individuals and groups can break through traditional boundaries and create truly innovative and groundbreaking work. As I read this on twitter (authentic source unknown)- “Even 100 of world’s best candle maker could not make an electric bulb.” The statement is strong enough to underline the value of idea, and how unconventional and unbiased approaches could change the world.
The statement also advance that biases can limit the scope of what is considered possible or valuable in each field and can lead to a lack of diversity and innovation. It is important for researchers and evaluators to be aware of the potential for ascertainment biases and to take steps to mitigate or avoid them in their work- at least in the professions where creativity, ideation and innovation have the central role.
Avoiding ascertainment biases in R&D project design:
There are several strategies that scientists can use to avoid ascertainment biases in their research:
Use diverse sample sets: One of the best ways to avoid ascertainment biases is to include a diverse sample set in the study. This means recruiting participants/samples from a variety of backgrounds, ages, conditions and demographics, to ensure that the study is representative of the population being studied.
Use multiple sources of data: Scientists can also avoid ascertainment biases by using multiple sources of data to support their conclusions. This may involve using a combination of qualitative and quantitative data or gathering data from different populations or locations.
Be aware of personal biases: This is suicidal biases. As we grow, we like to align ourselves towards certain ideas and ideologies, including in science. Scientists should also be aware of their own personal biases and how these biases may influence their research. This includes biases related to idea, competition, hypothesis, and workflows that often show up in a study design.
Use blind or double-blind studies: Blind or double-blind studies can help to avoid ascertainment biases by minimizing the influence of researcher biases. In a blind study, participants are not told which group they are in (e.g., treatment vs. control), while in a double-blind study, neither the participants nor the researchers know which group is which.
Conduct a systematic review: Finally, scientists can avoid ascertainment biases by conducting a systematic review of the literature. This involves gathering all relevant studies on a particular topic and analyzing them in a systematic way, to ensure that all relevant data is considered.
Get an evaluation: People in same organization feel associated with the project, and are directly affected by financial, social, academic success of it. The favorable project outcomes could positively influence their standing in the field. These are the ‘seed’ points of biases. Having a professional and independent review of assay designs could help improve whole project. Hire an independent consultant.
