What does research on implicit bias aim to solve? : 'Implicit' and 'Explicit'

In modern society, people do not openly express discriminatory attitudes as much as they used to. Many people believe they are egalitarians, for example, that they do not discriminate against others based on race. However, people still hold some attitudes or stereotypes towards social groups implicitly. In other words, implicit biases can differ from a person’s conscious beliefs and values. It is known that the values of implicit biases measured by the IAT are correlated with discriminatory behaviors.

I have organized the measurement methods and classification of results in this table.  In this context, “explicit” refers to what people express through responses to questions or checklists, while “implicit” refers to what is measured by the IAT.  The IAT scores are graded, but for simplicity, they have been divided into two categories.

Let’s consider cases where attitudes or stereotypes towards social groups or individuals are inconsistent between explicit and implicit measures.

When someone explicitly expresses attitudes or stereotypes towards a social group or individual, but these are not measured implicitly (EO), it may be due to social norms or situational pressures. For example, someone might act as a fan of a sports team in a region where it is popular, even if they are indifferent to it implicitly.

Conversely, when someone does not explicitly express attitudes or stereotypes but they are measured implicitly (OI), it may be due to social norms that discourage such expressions. For instance, someone might not openly express racist attitudes because it is socially unacceptable, but implicit biases are still measured. Alternatively, someone might believe they are not racist, yet implicit biases are detected.

As mentioned in the above examples, when something is not explicit, meaning it is not expressed by a person, it can be divided into cases where the person deliberately does not express it and cases where there is nothing to express according to their will or beliefs. When a person deliberately does not express something but it is observed implicitly (OI), it can become explicit (EI) depending on the situation, such as when expression is permitted. Similarly, a state (EI) can also become a state (OI). This applies to the combinations of state (EO) and state (OO) as well.

In summary, the reliability of explicit data is debatable due to factors like self-deception and impression management. Self-deception refers to falsehoods that the person themselves may not recognize as lies, while impression management involves strategies to be viewed more favorably by others.

Many people find the revelation of implicit biases (OI) uncomfortable. For those who consciously avoid expressing such biases, the exposure can be unpleasant, and for those whose implicit biases contradict their beliefs, it can be hard to accept. Those who are discriminated against may feel hurt and anxious.

From now on, I would like to examine how implicit biases lead to discriminatory behaviors.

Reference:

Blindspot: Hidden Biases of Good People: Banaji, Mahzarin R., Greenwald, Anthony G.: 8601300327310: Amazon.com: Books

https://banaji.sites.fas.harvard.edu/research/publications/articles/2017_Greenwald_AP.pdf

https://x.com/Channel4/status/1275861246726696960　

Implicit Bias and Prejudice: On the Translation of Implicit/Unconscious bias

In Japanese, implicit/unconscious bias is sometimes translated as ‘無意識の偏見’ (muishiki no henken, unconscious prejudice).  However, I would like to argue that it is inappropriate.

In 2009, Banaji and Greenwald, the proponents of “implicit bias,” used a statistical method called meta-analysis on 184 papers, revealing that the race IAT predicts racially discriminatory behavior. Subsequent studies have supported similar conclusions.

However, they state that the results of the race IAT do not imply prejudice.

For psychologists, prejudice generally refers to attitudes towards others based on group attributes (such as gender, age, race, nationality, occupation, religion, place of origin, hobbies, etc.), often negative. In a narrow sense, it refers to a person’s explicit and conscious dislike or aversion towards a specific group of people.

On the other hand, the implicit level of bias revealed by the race IAT is “It is not at all the same thing as racism or sexism, but it indeed is evidence of an association in our heads that I would call the roots of prejudice” Can we unlearn implicit biases? With Mahzarin Banaji, PhD (apa.org)

In other words, the racist behaviors clearly demonstrated in the paper that “the race IAT predicts racially discriminatory behavior” were “social behaviors in interracial interviews, doctors' treatment recommendations for a cardiac patient, and evaluation of job applications in a hiring situation” and were not the negative or hostile behaviors typically characterized as prejudice.

According to Banaji et al., translating Implicit Bias as “Implicit Prejudice” can be misleading. The term “prejudice” is often used negatively, potentially giving an unnecessarily negative impression of the term Implicit Bias. Implicit Bias can also be measured for positive things. Therefore, it is safer not to adopt the term “prejudice.” Since Bias also means “tendency,” it is better to use the term “バイアス” as it sounds in English.

"unconscious prejudice"

Reference:

Amazon.co.jp: Blindspot: Hidden Biases of Good People : Banaji, Mahzarin R., Greenwald, Anthony G.: Foreign Language Books

In Japanese: 

日本アンコンシャス・バイアス研究会: 潜在的バイアスと、偏見：「無意識の偏見」という訳について (bias-research.blogspot.com)

Unconscious bias, is it a general term for biases?

No, I believe it is inappropriate to handle it that way from various perspectives.

There are many psychological terms with the word “bias” besides latent bias. For example, there are confirmation bias and normalcy bias. Confirmation bias refers to “the tendency to collect and emphasize evidence that supports one’s beliefs or hypotheses, while ignoring evidence that contradicts them.” Normalcy bias refers to “the psychological tendency to regard certain levels of abnormality as within the normal range, such as thinking ‘it won’t be a big deal’ or ‘I’ll be fine’ in the face of disasters or risks.”

As mentioned earlier, implicit bias usually refers to implicit attitudes and stereotypes that influence our perceptions, judgments, and behaviors, often contradicting our explicit beliefs and values. Implicit bias is not treated as a general term for various biases, such as confirmation bias or normalcy bias, which are referred to as cognitive biases.

Given the history of the term “unconscious” being replaced by “implicit,” it is appropriate to list “unconscious bias” alongside “implicit bias.” Therefore, it is also appropriate not to treat unconscious bias as a general term for biases.

A search for peer-reviewed articles using “unconscious bias” in PsycArticles revealed that “unconscious bias” was used synonymously with “implicit bias,” supporting the above statement. (IJapan Society for Implicit Bias Research: American Psychological Association (APA) Database search: Unconscious < Implicit (jp-society-for-implicit-bias-research.blogspot.com)).

Additionally, a search on Google Scholar for “unconscious bias” and “implicit bias” and an examination of the top 10 influential papers for each term showed that both unconscious bias and implicit bias were interpreted similarly, further supporting the above statement (Japan Society for Implicit Bias Research: Public data including general publications: Unconscious < Implicit (jp-society-for-implicit-bias-research.blogspot.com)).

In Japanese: 日本アンコンシャス・バイアス研究会: アンコンシャス・バイアス、それはバイアスの総称？？ (bias-research.blogspot.com)

The Implicit Association Test (IAT)

"In line with the basic tenets of theories of associative learning and representation, the IAT rests on the assumption that it ought to be easier to make the same behavioral response (a key press) to concepts that are strongly associated than to concepts that are weakly associated (Greenwald et al., 1998)." NGB2005.PSPB.pdf (washington.edu)

Here is some examples.

In this example, the items to be categorized are displayed at the top of the screen, with “female” and “home” on the left side and “male” and “career” on the right side. The participant sees the word “she” displayed in the middle and categorizes it to the left side. The categorization is done by pressing the keyboard keys associated with the left or right side (for example, 'e' for left and 'i' for right). The words displayed in the middle include “he,” “man,” “boy,” “she,” as well as “kitchen,” “office,” “garden,” “profession,” and “marriage.”

Next, the categories are switched, with “career” and “home” reversed. The participant sees the word “office” displayed in the middle and categorizes it to the left side, but many people may take longer to respond compared to the previous categorization. This test measures whether we hold implicit stereotypes that associate women with home and men with career. Importantly, even female participants may strongly associate women with home.

The IAT can also measure implicit “attitudes.” Attitudes refer to a person’s likes or dislikes, such as liking flowers or disliking insects.

In this example, the categories “flowers” and “good” are on the right side, while “insects” and “bad” are on the left side. The words displayed in the middle include “lily,” “evil,” “gentle,” “flea,” “rose,” and “heaven.”

When the categories “pleasant” and “unpleasant” are switched in the test, participants (especially those who do not particularly like insects) will likely take longer to categorize compared to the previous classification. Of course, the opposite may occur for those who love insects.

The IAT’s ability to measure implicit stereotypes and attitudes has had a significant impact on the academic community, leading to numerous related studies and publications. The IAT can be taken on Harvard University’s “Project Implicit” website (Project Implicit (harvard.edu)). To date, the IAT has been taken at least 40 million times. 

The term “Implicit Bias” is more commonly used and recognized globally than “Unconscious Bias,” likely because this test is so well-known.

Understanding what this test aims to measure and how it does so reveals that it is fundamentally different from the “unconscious bias surveys” in Japan, which use question formats like “Do you think women are suited to be leaders?”

Greenwald and Banaji, the proponents of unconscious bias, often referred to their early research memories during conferences and academic gatherings. Memory is classified into implicit memory and explicit memory based on its nature. The terms implicit and explicit are used not conceptually but based on research experience. Implicit refers to indirectly measured results, while explicit refers to directly measured results. For example, when stimuli are given to amnesiac patients who cannot remember new information, the effects may manifest in their actions even if they are not consciously aware of it. Such memories are observed indirectly and are called implicit memories.

However, it is stated that implicit measurement is not necessarily unconscious measurement, and explicit measurement is not necessarily conscious measurement. Research results suggest that both direct and indirect measurements need to be assumed to combine the influences of conscious and unconscious mental processes.

Greenwald and others named the test that measures stereotypes and attitudes indirectly as the Implicit Association Test (IAT).

If we follow their naming intention, it is preferable not to list implicit bias alongside unconscious bias. It is desirable to use the term implicit bias for implict stereotypes and implicit attitudes measured by the IAT.

References:

About the IAT – Project Implicit

ethz.ch/content/dam/ethz/associates/services/Anstellung-Arbeiten/diversity/Gender und Stereotype/Auswertung IAT final_RS_EN.pdf

Can we unlearn implicit biases? With Mahzarin Banaji, PhD (apa.org)

Blindspot: Mahzarin R. Banaji, Anthony G. Greenwald, Eric Martin: 9781491528860: Amazon.com: Books
https://banaji.sites.fas.harvard.edu/research/publications/articles/2017_Greenwald_AP.pdf

https://www.larryjacoby.ca/images/a%20process%20dissociation%20framework.pdf

"Unconscious Bias" in Japan, "Implicit Bias" in World / U.S.

 I checked the search volumes for the following terms on Google Trends. The horizontal axis represents the period from January 2004 to August 2024 (as of August 23, 2024), and the vertical axis shows the percentage relative to the highest search volume during that period, set at 100%. If the text appears small and hard to read, please click on the image.

In Japan, as shown in the above figure, the term “アンコンシャス　バイアス” (orange, Unconscious Bias) is searched more frequently than “潜在的　バイアス” (dark blue, Implicit Bias) (as of August 23, 2024). 

Looking at the graph, the number of searches for “アンコンシャス　バイアス” has significantly increased since 2018. I have considered the reasons for this.

The earliest introduction of the term “アンコンシャス　バイアス” in Japan seems to be around 2016(essay.pdf (djrenrakukai.org)). The results collected by Hisako Otsubo were compiled into a proposal in English (提案書（表紙） (oist.jp)) in 2016. This proposal listed “implicit/unconscious bias,” as is often seen in English-speaking countries.

Subsequently, in 2017, a leaflet titled “Do you know about unconscious bias?” (UnconsciousBias_leaflet.pdf (djrenrakukai.org)) was created by Otsubo and Machi Dilworth and distributed to universities, academic societies, and the Cabinet Office (first edition in 2017, revised edition in 2019).

In this leaflet, the term “Unconscious Bias” is chosen instead of “Implicit Bias.” Additionally, the leaflet uses the same font size for “Unconscious Bias” and “無意識のバイアス,” suggesting that people are expected to read it as “Do you know about Unconscious Bias?” Furthermore, this leaflet is published as a PDF file named “UnconsciousBias_leaflet.pdf.” Therefore, the leaflet contains multiple elements that promote the awareness of “Unconscious Bias” rather than “Implicit Bias.”

Thus, it is speculated that the distribution of this leaflet since 2017 is one of the reasons why “Unconscious Bias” has been significantly more searched on Google in Japan since 2018, compared to “Implicit Bias.”

Why was “Unconscious Bias” adopted in this leaflet instead of “Implicit Bias”?

One reason could be that in 2014, Google held a seminar titled “Unconscious Bias @ Work” (Unconscious Bias @ Work | Google Ventures (youtube.com)), which became a topic of discussion. This video has been viewed over 370,000 times as of October 7, 2024. The reason for holding this seminar was due to external criticism that more than half of the people featured in Google Doodles (temporary changes to Google’s homepage to commemorate notable individuals’ birthdays, etc.) were white men, indicating a need for improvement. CNN also reported on this criticism, confirming that it was a topic of discussion at the time (Google Doodles team makes strides toward diversity pledge | CNN).

Adopting “Unconscious Bias” in this leaflet is speculated to have laid the groundwork for misunderstandings and misuse, as will be introduced later.

So, what happens if a similar survey is conducted worldwide?

I checked the Google search volume for the following terms on Google Trends. The horizontal axis represents the period from January 2004 to October 2024 (as of October 12, 2024), and the vertical axis shows the percentage of the maximum search volume during that period, with the highest search volume set at 100%. If the text is small and hard to read, please click on the image to enlarge it.

Globally, “Implicit Bias” is searched more frequently than “Unconscious Bias.” As we have seen, this is likely because “Implicit Bias” is more prevalent in academic circles. Additionally, the search volumes for both “Implicit Bias” and “Unconscious Bias” have increased since 2014. It can be said that 2014 was the year when society began to show interest in this issue.

For most of the period, the search volume for “Unconscious Bias” follows the trend of “Implicit Bias,” but in 2020, the search volume for “Implicit Bias” spikes significantly. Why is this?

In that year, African Americans George Floyd and Breonna Taylor were killed by police officers in the United States, leading to the intensification of the Black Lives Matter (BLM) movement across the country. At the same time, the term “Implicit Bias” became widely known (Taking a hard look at our implicit biases — Harvard Gazette). This might be reflected in the graph.

Bias and Police Killings of Black People : NPR

Additionally, the search volume for “Implicit Bias” also spiked in 2016. This spike appears to be unique to “Implicit Bias” and does not follow the general trend of “Unconscious Bias.” This is likely due to Hillary Clinton mentioning “Implicit Bias” during the U.S. presidential debate that year. At that time, the issue of police shootings of African Americans was already a significant concern. Clinton stated that “Implicit Bias” is not just a problem for police officers but for everyone, and she intended to address it. This debate was watched by 83 million people. The impact of this event might be reflected in the Google Trends data.

As mentioned above, the term “Implicit Bias” is globally recognized as the mainstream term.

From here, I would like to introduce Unconscious Bias/Implicit Bias in more detail and provide foundational knowledge to illustrate its misuse in Japan. First, how is Unconscious Bias/Implicit Bias measured?

Public data including general publications: Unconscious < Implicit

First, I confirmed that the term ‘unconscious cognition’ has been academically replaced by the term ‘implicit cognition.’ Then, I examined ‘unconscious bias’ and ‘implicit bias‘.

The tool is the Google Books Ngram Viewer. The Google Books Ngram Viewer displays how often the inputted Ngram appears in the corpus of books (a large text dataset collected and organized for natural language processing) over the selected years. An Ngram is a sequence of N items in a text or speech, for example, ‘kindergarten’ is a 1-gram, and ‘nursery school’ is a 2-gram.

The figure below shows the search results for ‘implicit cognition’ and ‘unconscious cognition’ (as of Octorber 18, 2024) The vertical axis indicates the respective proportions of these two search terms appearing in 2-grams in books published in the United States between 1800 and 2022, digitized by Google and written in English. The settings are case-insensitive. Although the text is small, the data for ‘implicit cognition’ is represented in blue, and ‘unconscious cognition’ is represented in red. The vertical dashed line represents the year 2000.  Click to enlarge the image.

It can be observed that the usage frequency of “Implicit Cognition” has increased more than “Unconscious Cognition” since the 1990s. This trend is believed to be influenced by the developments in scientific psychology described earlier by Greenwald and Banaji. Additionally, both terms have seen an increase in usage since 2000, but according to 2022 data, “Implicit Cognition” is used approximately four times more frequently than “Unconscious Cognition.” Incidentally, the value of “Unconscious Cognition” rose between 1870 and 1890, but the reason for this is currently unknown (it predates Freud’s works).

The following graph includes the terms “implicit bias” and “unconscious bias” (as of Octorber 18, 2024). Green represents “implicit bias” and yellow represents “unconscious bias.” Both terms have seen an increase in usage since 2000, with a sharp rise after 2010. According to 2022 data, “Implicit Bias” is used approximately 1.6 times more frequently than “Unconscious Bias.” It can be seen that “Implicit Bias” and “Unconscious Bias” are used more frequently than “Implicit Cognition” and “Unconscious Cognition.”  Click to enlarge the image.

Next, I examined the influence of academic literature using these terms by looking at the number of citations and the publication year of documents searched on Google Scholar. I searched for “implicit bias,” “unconscious bias,” “implicit cognition,” and “unconscious cognition” on Google Scholar and sorted the top 100 results for each term by a measure of influence (citations per year since publication) as of August 17, 2024. Note that Google Scholar does not distinguish between uppercase and lowercase letters. The following graph illustrates this data. Please note that the vertical axis is logarithmic. From this graph, it can be seen that papers using the terms “Implicit Cognition” and “Implicit Bias” have a greater influence than those using “Unconscious Cognition” and “Unconscious Bias.”

Among these, the top two papers with outstanding values for (citations per year since publication) are those searched under “Implicit Cognition.” These are the papers previously mentioned, published by Greenwald and Banaji in 1995, and by Greenwald et al. in 1998. The vertical axis is logarithmic, and the influence of the top-ranking papers is literally orders of magnitude greater. These papers suggest that the term “unconscious” was likely replaced by “implicit” in the 1990s.

Below is a graph plotting the top 100 results by publication year and number of citations. Note that the vertical axis is also logarithmic. The search results for “Implicit Cognition” and “Unconscious Cognition” span from the 1980s to recent years, while those for “Implicit Bias” and “Unconscious Bias” do not appear before 1999. It can be seen that the term “Implicit” is used in more frequently cited literature compared to “Unconscious.”

From the data we have seen so far, it is clear that “Implicit Cognition” and “Unconscious Cognition” started being used first, followed by “Implicit Bias” and “Unconscious Bias.” The current usage (estimated from the graphs of period and frequency of use, or period and number of citations) suggests that “Implicit Cognition” and “Implicit Bias” are used more frequently than “Unconscious Cognition” and “Unconscious Bias.”

The following are the results of extracting the definitions of the top 10 papers for “Unconscious Bias” and “Implicit Bias” based on citations per year since publication, summarizing them with ChatGPT(note: two papers related to deep learning and machine learning in the results for “Implicit Bias” were excluded.  ChatGPT used 2024/10/18).

Implicit bias refers to unconscious attitudes or stereotypes that influence our understanding, actions, and decisions. These biases can lead to behaviors that diverge from an individual's explicit beliefs and intentions, even among those with the best intentions. They are pervasive, affect how we interact with others, and can contribute to discrimination, particularly in contexts like healthcare, where providers may hold negative biases against marginalized groups. Implicit biases operate outside of conscious awareness, complicating efforts to address discrimination, as they challenge the assumption that people act solely based on their consciously held beliefs.

Unconscious bias refers to the attitudes or stereotypes that unconsciously influence our perceptions and decision-making, affecting behavior and interactions without our awareness. These biases can be both favorable and unfavorable, impacting various contexts, including workplaces and healthcare. Despite intentions to act fairly, individuals often harbor implicit biases that can lead to discriminatory behaviors, such as unequal treatment in medical diagnoses or biased performance evaluations in corporate settings.

Unconscious bias training aims to raise awareness about these biases and their effects, yet its effectiveness remains debated. Critics argue that knowing about biases doesn't guarantee behavior change, especially when structural constraints are present. The training has gained popularity in organizations as part of diversity and inclusion initiatives, but it is suggested that a more profound, long-term approach may be necessary to create meaningful change. Ultimately, unconscious biases are powerful because they operate outside of conscious awareness, often resulting in actions that contradict one's explicit beliefs.

This is a rough method, but I hypothesize that in influential literature found through Google Scholar, ‘unconscious bias’ and ‘implicit bias’ are considered synonymous.

Upon individually reviewing each interpretation, one paper mentioned apophenia (the tendency to perceive meaningful connections between unrelated things) as an example of unconscious bias. However, the author believes that the other literature generally aligns well with the above summary.

From previous investigations using the American Psychological Association’s database and public data, it appears that ‘unconscious bias’ and ‘implicit bias’ are used interchangeably, which is why they are often listed together. Additionally, it was found that ‘implicit bias’ is more mainstream than ‘unconscious bias.’

For reference, when the Japanese term ‘アンコンシャス・バイアス’ (unconscious bias) was similarly checked as a search term in Google Scholar, the citation count/year since publication was 1 for 5 cases, and zero for the remaining 95 cases. This indicates that the number of influential and well-cited documents written in Japanese about unconscious bias is very small.

Next, I would like to introduce the background of how the term ‘unconscious bias’ was introduced in Japan.

In Japanese: 日本アンコンシャス・バイアス研究会: 一般的書を含む公開データで見る、アンコンシャス・バイアス　＜　潜在的バイアス (bias-research.blogspot.com)

American Psychological Association (APA) Database search: Unconscious < Implicit

 Japanese site:

日本アンコンシャス・バイアス研究会: アメリカ心理学会のデータベースで調べてみた (bias-research.blogspot.com)

Using APA PsycNet database, the above words are the search queries for peer-reviewed journal articles since 1950 to 2023.

According to Banaji and Greenwald, 

"A quarter century ago, the word "unconscious" -having fallen out of favor in scientific psychology earlier in the twentieth century - was barely to be found in the scientific journals that we read and in which we published our research.  (...) the term "unconscious cognition" (...) was surpassed in the 1990s by the related term "implicit cognition"."

 Blindspot: Hidden Biases of Good People

You’ll see in the figure above that the use of “unconscious” has remained relatively low compared to “implicit,” which began to surpass “unconscious” in the 1990s.

It is also evident that “implicit bias” is used more frequently than “unconscious bias.”

So, are the terms “implicit bias” and “unconscious bias” used in the same way?

In conclusion, they are used in the same way.

A search of PsycArticles for peer-reviewed papers using “unconscious bias” since 1950 confirmed that “unconscious bias” is used in the same sense as “implicit bias.” Academically, the use of “unconscious bias” is in the minority, and when it is used, it is understood to mean “implicit bias.”

Although scientific psychologists prefer to use the term “implicit” rather than “unconscious,” the two terms have been used almost synonymously. However, a decisive event solidified the prominence of “implicit.” This was the publication of papers by Greenwald and Banaji in 1995 and by Greenwald et al. in 1998. These papers had a significant impact, with over 10,000 and 17,000 citations respectively as of August 17, 2024.

Since then, it can be considered that, as seen in data from the American Psychological Association, the number of papers using “implicit” has surpassed those using “unconscious.”

Just for your reference, the reason of "unconscious -having fallen out of favor in scientific psychology earlier in the twentieth century" is:

"in part a reaction against the great popularity of Sigmund Freud and psychoanalytic theory and in part also a reflection of American behaviorists' eschewal of both conscious and unconscious mentality in much of the 20th century"

2017_Greenwald_AP.pdf (harvard.edu)

The more specific factor behind the backlash of popularity was Freud's arguments had detached as they have remained from scientific varification.  

"Then it have a greatly reduced impact on scientific understanding of unconsdious mental life."  Blindspot: Hidden Biases of Good People

In summary, scientific psychologists have been using “implicit” over time. “Implicit bias” and “unconscious bias” have the same meaning, and this has been verified through the APA database.

https://banaji.sites.fas.harvard.edu/research/publications/articles/2017_Greenwald_AP.pdf

Blindspot: Hidden Biases of Good People

https://faculty.washington.edu/agg/pdf/Greenwald_Banaji_PsychRev_1995.OCR.pdf

https://faculty.washington.edu/agg/pdf/Gwald_McGh_Schw_JPSP_1998.OCR.pdf

https://www.researchgate.net/publication/374093138_Awareness_of_implicit_attitudes_Large-scale_investigations_of_mechanism_and_scope

What is the Unconscious Bias?

An example of unconscious bias affecting people’s perceptions is the story of blind auditions. In the 1970s, less than 10% of the members of a renowned American orchestra were women. However, when a screen was placed between the performers and the judges during auditions, the percentage of women increased to nearly 40% (“Blindspot: Hidden Biases of Good People” by M.R. Banaji and A.G. Greenwald). If you asked the judges, “Do you think men are better performers?” they would likely respond, “No, I believe there is no gender difference in musical ability.” However, when the gender of the performers was apparent, the judges were influenced by it.

Unconscious bias is often mentioned alongside implicit bias.  The terms used interchangeably.

https://en.wikipedia.org/wiki/Cognitive_bias

https://www.oxfordlearnersdictionaries.com/definition/english/implicit-bias?q=implicit+bias

https://www.simplypsychology.org/implicit-bias.html

Let me explain about Implicit Bias.

It is said;

"Implicit biases are discriminatory biases based on implicit attitudes or implicit stereotypes. Implicit biases are especially intriguing, and also es-pecially problematic, because they can produce behavior that diverges from a person’s avowed or endorsed beliefs or principles. The very existence of implicit bias poses a challenge to legal theory and practice, because dis-crimination doctrine is premised on the assumption that, barring insanity or mental incompetence, human actors are guided by their avowed (explicit) beliefs, attitudes, and intentions" ResearchGate

"hidden bias"

In Japanese:

日本アンコンシャス・バイアス研究会: アンコンシャス・バイアスって何？ (bias-research.blogspot.com) 

日本アンコンシャス・バイアス研究会: 英語では別名もあった！？「潜在的バイアス」 (bias-research.blogspot.com)

Unconscious Bias is Being Misused in Japan

Currently, the misuse of “unconscious bias” is spreading in Japan. As a result, the issues we originally intended to address are being neglected. Therefore, in this blog, I will explain the original meaning of the term, then discuss why its misuse is problematic. Following that, I will point out specific examples of misuse. Finally, I will propose solutions to this issue. 

"misunderstood of unconscious bias in japan" using japanese amination taste."

See in Japanese: 日本アンコンシャス・バイアス研究会: アンコンシャス・バイアスは、誤用されている (bias-research.blogspot.com)