Choosing the road less travelled: Radical kindness and human-centred test design

When I heard Dr Jenna Mittelmeier talk about radical kindness in higher education I did not have a road to Damascus moment, because I was already converted. I just didn’t have a term for my beliefs. What I experienced was an epiphany. Everything she said resonated (I was like a struck tuning fork) with what I feel personally and professionally about education.

Radical kindness means recognising the humanity of students, seeing them as individuals, and treating them accordingly at every stage of their education.* All this may appear, to some, as rather saccharine and impractical. I disagree. My colleagues and I work every day to put the principles of radical kindness into practice through the holistic, human-centred test design of LANGUAGECERT Academic.

Before I show how these principles become question papers, I need to take a detour and talk about a concept that is central to the discussion of test design.

The concept which must not be named

All stand-up comics have horror stories about losing the room. The moment when they look at their audience and see people yawning, checking their phones or whispering to each other. I cannot remember how many presentations about language testing I have given over the years. Thankfully, unlike comedians, I have never been heckled but I have seen eyes glaze over, heard the involuntary sighs and watched the opening of laptops when I mention the unmentionable: test construct. The very antithesis of audience engagement and assessment literacy kryptonite.

What is a construct and why does it provoke such a response whenever it is mentioned? I think the adverse reaction is caused by gnomic definitions, to the uninitiated, such as this: construct – what a test measures. This is far too get cape, wear cape, fly. Explaining the tricky bit between wearing and flying gets nebulous and circular. The concept becomes a modifier for a range of language testing terminology: construct validity, construct variance and so on and on and ...

The curse of knowledge

With the above I have probably provoked the ire of my fellow assessment professionals and before I am excommunicated I need to appease them. The problem with explaining a construct is what Steven Pinker calls ‘the curse of knowledge’. Experts are experts (I am not claiming to be one!) because their level of knowledge and understanding in their field is by its nature extraordinary when compared to the knowledge of those not in the field.

"When we know something well, we don't realise how abstractly we think about it. And we forget that other people, who have lived their own lives, have not gone through our idiosyncratic histories of abstractification.”**

My personal abstractification and therefore understanding of construct is grounded in questions and choices.

Asking the questions

All test developers ask themselves why, who and what am I testing? The process is iterative because why, who and what not only relate to test takers but also to stakeholders such as higher education, governments and other recognising organisations that accept a test’s surrender value. Why does a university or college need an admissions test, who are they accepting and what does the institution want to be tested? Questions overlap and combine into complex permutations.

From the answers to all these questions the test developer derives the ultimate purpose of the test and what to measure, which brings us to the construct (what a test measures). The LANGUAGECERT Academic construct is English communicative ability and language proficiency for an academic environment.

Job done? Of course not. If only it were that simple. Questions of how we test and measure this construct must be answered and the right choices made.

Making the right choices

One can choose to design, produce and deliver a test which puts convenience, for the assessment body, before test takers’ real-world language needs. One can choose to offer a test where the delivery platform dictates the content and puts the needs of the platform before the needs of the people taking the test. One can choose to design, produce and deliver a complete test of academic English, across all four skills, with every item assessing communicative ability. Or choose not to. 

When choices must be made where roads diverge, our principles of holistic, human-centred test design acts as a compass to show the correct way along the road less travelled by. And that makes all the difference. 

We think of and treat our test takers as individuals and not candidate numbers. We aim to enable our test takers to handle the academic rigour of studying and actively participate in all aspects of their higher education experience. We want our test takers to be heroes – to get cape, wear cape, fly. 

*This same principle applies to teachers, lecturers and administrators. Also, in the context of assessment it applies to everyone involved in the end-to-end assessment process, those who write, administer and mark the tests.

**The Sense of Style. The Thinking Person’s Guide to Writing in the 21st Century, Steven Pinker.