My unique method for testing speaking ability in the EFL classroom

• The class most often taught by native English speakers…
• Typical methods of testing speaking ability in EFL settings
• Problems with these methods for testing speaking ability in the EFL classroom
• How I found a better method for testing speaking ability in the EFL classroom
• Getting down to business—preparing for the test
• Tailoring my unique method for testing speaking ability in the EFL classroom
• There was one vital difference to my method for testing speaking ability in the EFL classroom
• The pros of my method for testing speaking ability in the EFL classroom
• You’ve got to be prepared for the unexpected in every conversation
• No one expects the Spanish Inquisition
• Cons of my method for testing speaking ability in the EFL classroom
• Any time partners are involved, one inevitably does the heavy lifting
• We often hold non-native speakers to higher standards—but not always
• What do you think of my method for testing speaking ability in the EFL classroom?

I taught English at the university level for 17 years. I assessed my students’ abilities via midterm and final tests. During my time, I taught several types of classes.

• Listening
• Writing
• Pronunciation
• Presentation
• General English

The class most often taught by native English speakers…

But the class I taught most often was the infamous conversation class. This is the class commonly taught by native English speakers. Sometimes called “speaking class,” the goal is to get students speaking in English. As student levels vary, so will the complexity of the conversations. Lower-level students may discuss hobbies and preferences. More advanced students could discuss the moral and ethical implications of assisted suicide.

Administering midterm or final exams for these classes can be challenging. What is the “best” way to assess someone’s English conversational ability? In writing class, it was pretty easy to create a written test to assess if my students had learned what I’d taught them. In presentation classes, presenting in front of the class was an effective evaluation. But what’s the best method for testing speaking ability in an EFL classroom?

Typical methods of testing speaking ability in EFL settings

I’ve seen instructors give written tests for speaking classes. I think I did the same for my General English classes initially. But that’s not the most effective way to assess someone’s English-speaking ability. Writing answers or creating a dialogue allows for evaluating comprehension and grammar.

But that isn’t an accurate reflection of one’s conversational skills. Conversation involves listening, comprehension, processing and speaking—all within a very short time. 

You have time to assess and understand the question on a written test. There’s time to write your answer, erase it, fix it, and perfect it several times before submitting your paper. You don’t have the same kind of pressure as when someone has asked you a question and is waiting for your answer.

Most instructors I knew conducted speaking tests using one of these methods:

• Questions and Answers—Students were given a list of 50 questions. They were expected to be able to answer them using proper grammar. The instructor would ask them ten or so questions, perhaps with follow-up questions.
• Speeches—Students memorized a speech on a particular topic to deliver to the instructor.
• Conversation with the instructor—Students had conversations with their instructor on pre-determined topics.
• Role Play—Students assumed roles (i.e., waiter & customer) and delivered a dialogue as those characters.

Problems with these methods for testing speaking ability in the EFL classroom

All these methods, at the very least, encourage students to speak. The instructor can assess pronunciation, intonation, comprehension, and grammar. But they all have their downsides.

• Q&A—Memorizing answers to questions doesn’t necessarily translate to speaking ability. Usually, there is no chance for the students to ask questions themselves.
• Speeches—Technically speaking, this is another memorized format. It doesn’t accurately assess a student’s speaking ability. It’s devoid of interaction, a crucial part of conversation.
• Conversation with the instructor—As a native speaker, the instructor tends to dominate and steer the conversation. They ask follow-up questions to elicit more information. There’s also a danger of the instructor speaking too long when a student asks a question. Students may ask a series of follow-up questions to keep the instructor talking.
• Role Plays—This can feel too unnatural, almost like acting. Being asked to assume the role of a waitperson or police officer can feel very awkward. It doesn’t give students a chance to speak using their own experiences.

While all better than a written test, I felt there must be a better way.

How I found a better method for testing speaking ability in the EFL classroom

In 2009, I started my master’s degree in TESOL (Teaching English to Speakers of Other Languages). I studied at Sookmyung Women’s University—only the undergrad program is female-only 😉. The prerequisite to that course was a 4-month practical TESOL course.

The majority of students were Korean. I’m not sure how many classes there were, but in my class of 16, I was one of 4 non-Korean students. The Koreans in the class were the highest-level English speakers. The other classes consisted of Koreans or non-native English speakers.

Our second language acquisition (SLA) class midterm involved a conversation with our partner. We were given 3-4 scenarios. One of the scenarios was as follows:

You are setting up a new program at your school. Use the theories and learnings from our SLA course to create a curriculum for a language course. Discuss elements to include in your curriculum to put those theories into practice.

This was a way to assess the Korean students’ speaking ability and course knowledge. For the native speakers, we had to show we understood the course content. With four scenarios to prepare, creating and memorizing four different dialogues was impossible.

We were shown a recorded demonstration by two native speakers of what this test might look like. We were only shown it once, and the goal was not to replicate or parrot the conversation. It was simply an example of what was expected. Besides, there were three other possible scenarios to prepare for.

Getting down to business—preparing for the test

My partner (another native English speaker) and I took the test seriously. We held study meetings where we highlighted key areas and concepts, and brainstormed topics we needed to cover. We agreed on who’d speak on which topics based on who felt most comfortable with which concepts.

On test day, we were quietly confident that we would do well. But having never taken a test like this, there were some nerves. When we were called into the testing room, the instructor gave us a pleasant surprise. He informed us we could choose which topic we’d like to do.

Though well prepared for all the topics, we were most confident with the one on curriculum design. Our instructor sighed and mentioned that was the topic most people had chosen. But since he’d given everyone else the choice, he told us to begin. We did our best. It wasn’t perfect, but we thought we’d covered most of what we’d decided to talk about.

I’ll never forget the moment when we finished. Our instructor put his pen down, put his head in his hands and kind of mumbled, “Jesus Christ, guys!” What had we done wrong? Had we misunderstood the assignment? I didn’t know how, but I was sure we’d tanked our test.

My heart resumed beating with his following sentence. “Guys, that was even better than the demonstration. I wish I’d recorded that.” Phew!

Needless to say, we both aced that test—but we’d put the work in. We didn’t have to worry about our English grammar while speaking. But the experience gave me an idea.

Tailoring my unique method for testing speaking ability in the EFL classroom

Shortly after that, I adapted the format for my conversation class tests. In my classes, I did a lot of partner and group work. I focused on communicating in English—you know, having conversations.

My students already sat with different partners each day. I did this by generating a random seating plan for each class. This simulated the awkwardness of talking with someone you didn’t know well. Practicing speaking during class seemed like the perfect preparation for these speaking tests.

Two weeks before a test, I’d give the students five conversation topics based on things we’d covered in class. The scenarios defined situations, contexts, and expected outcomes. I’d go over the topics and my grading rubric. Then, I’d tell the students to come to the following week’s class ready to practice. For midterms, the conversations were three minutes long and four for finals. 

The week before the test, we’d do a “dry run” in class. I’d randomly select one of the five topics, start a three or four-minute timer, and say, “Go!” Once the timer went off, the students would change partners, and I’d choose another topic. I repeated this process until the students had practiced all five topics. 

There was one vital difference to my method for testing speaking ability in the EFL classroom

The key difference with my testing method was that I conducted my tests with random partners. Students didn’t know who their partner would be until moments before the test. I had a list, and I’d tell the first two on the list to join me in the testing room. This prevented students from writing and memorizing dialogues.

Here’s an example of a scenario I would give my students.

You meet your partner by chance on the subway. After some brief small talk, discuss plans for the weekend and agree to watch a movie together. Find a movie currently in theatres that you’d both enjoy. Decide on a time and place to meet. If there’s still time remaining, discuss what you might do together after the movie.

All the parts of this scenario had been covered in class during the semester. Students had time to prepare the topics. After practicing the topics, they still had a week to prepare for the test. They could adjust for any “surprises” that may have come up during the practice session. My students had time to look up the vocabulary necessary to communicate what they wanted to say. They could also review grammar patterns.

The pros of my method for testing speaking ability in the EFL classroom

I preferred this method because it was the best way I found to simulate a conversation. It wasn’t a memorized dialogue or a role play. Granted, it was still a “staged” scenario. But I wanted the students to be themselves during these conversations. The goal was to get them to speak about their experiences, preferences, and opinions. This is what they’d do in an actual conversation.

The above scenario allowed me to assess several things: 

• Natural greetings—I spent a lot of time on this in class. “How are you? Fine, thank you, and you?” evoked the same reaction as a Grand Open sign does now. 😉
• Small talk—Could students easily talk about school, the weather, etc.?
• Movies and actors—We spent some time studying movie genres in class. We also expressed likes and dislikes and talked about actors and actresses.
• Making plans—By agreeing on a time and place to meet, they’d demonstrate knowledge of another topic we’d covered in class.
• Comprehension—Did each student understand their partner’s questions and responses? If not, did they use conversational tools to help each other understand?
• General conversational ability—Could they keep the conversation going for the entire 3-4 minutes? If they completed the goal, could they keep the conversation going on more general topics?

The topic was defined enough that they knew what they’d have to talk about—but that was it. I’d warn students, “Don’t come to the test convinced you’re going to see the next Marvel movie. Your partner may hate superhero movies. You have to agree on something you both like.” Having random partners prevented students from writing dialogues they could memorize. They had to speak, listen, and react—you know, have a conversation.

You’ve got to be prepared for the unexpected in every conversation

As I explained to my students, memorizing a dialogue is unnatural—unless you’re an actor/actress. Every conversation is an unknown. You may have had the same conversation every day with your best friend for the past two years.

• A: How’s it going?
• B: Not bad, and you?
• A: Pretty good.

But tomorrow, when you ask, “How’s it going?” and your friend responds, “My dog died,” you have to react. “Pretty good” not only doesn’t work grammatically—it’s not an appropriate reaction. And that’s what I wanted to test with my students. 

I wanted to see how well they could process information and communicate in the moment. Could they use the tools I’d taught them for clarifying when they didn’t understand something? I wanted to see how they could communicate using English in real time.

No one expects the Spanish Inquisition

My favourite example of the “unknown” aspect of conversations happened during an early test. Two female students were discussing their plans for New Year’s Eve. Part of the conversation between them went something like this:

• A: So, what are you going to do on New Year’s Eve?
• B: I’m going to go to a hotel.
• A: Oh, with your family?
• B: No, with my boyfriend.
• A: Oh…..OH!

I deserve a medal for keeping a straight face during that one. FWIW, as a Canadian, I’ve taken several French conversation tests in my time. I never felt the need to inform my teacher/professor of my intention to have sex with my girlfriend during a French test.

Cons of my method for testing speaking ability in the EFL classroom

This was an unusual test style for my students, most of whom were used to (and preferred) rote memorization. It was not a test that one could cram for the night before. It wasn’t uncommon for students with good English speaking ability to do poorly on these tests.

This happened when they hadn’t reviewed the topics and understood the goal of the scenario. Such students would see “movies” for the scenario above and only talk about movies. They never got around to making plans to see a movie with their partner.

These tests were challenging to orchestrate. I needed two classrooms—a waiting room and a testing room. My classes often had upwards of 40 students. It wasn’t fair for the final pair to wait 90+ minutes for their turn. So, the day before the test, I’d post “groups” for the 2-hour block. For example, if the test was scheduled for 1:00-3:00 pm

• Group 1: Students 1–10 (1:00-1:30)
• Group 2: Students 11–20 (1:30-2:00)
• Group 3: Students 21–30 (2:00-2:30)
• Group 4: Students 31–40 (2:30-3:00)

Thus, no student had to wait more than 30 minutes for their turn.

It was a lot of pressure packed into a short time. Midterm/final tests often take an hour or two. Having 30-40% of your grade based on 3–4 minutes meant that if a student choked, they could really tank their score. But that was part of the test. Conversing in a foreign language is stressful. Speaking with a stranger and not knowing what they’ll say is particularly stressful.

Any time partners are involved, one inevitably does the heavy lifting

Stronger students sometimes had to “carry” the weaker students. But that was reflected in their grades. I knew who’d put the work in and who hadn’t. I knew who’d prepared and who had been lazy.

If one student kept trying to guide their partner, and their partner gave one-word answers, the first student got a higher score. I told my students I would always reward effort. Ability never hurt, but effort was always important to me. Spoken language is almost always imperfect.

The students who typically got the highest grades were the ones who had thoroughly prepared. They also worked with their random partner, aware that working together would likely benefit them both.

And though part of their grade was on grammar, I was more interested in the ability to communicate. Native speakers don’t always speak eloquently or with perfect grammar. Take this transcript, for example: 

Thank you very much. Thank you very much, everybody. It’s great to be here. What a place. Did you like the flyover? That was pretty good. We were going low and fast. We went. That was moving along. It was pretty low. 

And later on..

He is very angry. He was so angry the other night, he’s like seething. You ever see a guy make a speech, such anger and so uncomfortable, just shouting, but he’s terrible. He’s just terrible. Should have never been there, we all know that, should have never been there.

Notice the simple language, incomplete sentences, and disjointed train of thought. It would be easy to assume this was a second-language speaker struggling with English. But you’d be wrong. It’s a transcript from a recent Donald Trump rally. 

We often hold non-native speakers to higher standards—but not always

I’ve written before about how non-native speakers are held to a higher standard. I didn’t ignore grammar on tests, but I considered other aspects of speaking.

As I explained to my students, language is about communication. I would ask them, “Outside of this classroom, in the ‘real world,’ has anyone ever corrected your English? What happens if you check into a hotel overseas and say, ‘We is staying for two night.’”

Does the clerk reply, “Do you mean you ARE staying for two nightS?” Of course not! Language is about communication. If you can exchange information and get the keys to your room, then mission accomplished!

What do you think of my method for testing speaking ability in the EFL classroom?

I shared this testing method with fellow instructors, both those I worked with and those I didn’t. I’m unaware of anyone else using this method. It was more effort on my part, but I felt it was the best way to evaluate my students’ conversational abilities. 

I always recorded tests for the rare occasion a student would dispute a grade. I estimate this happened fewer than five times during my 17 years of teaching. Recording enabled me to replay the conversation and show the student why I’d given them the grade. This was usually enough to quell any protests.

So, what do you think—would you consider using this method in your conversation classes? Have you got a better way to test English conversation ability? What holes do you see in the approach I used? I’m keen to hear your thoughts in the comments.

Subscribe to this blog or follow me on Instagram, Threads, Facebook, or LinkedIn for more content and updates.