Here is the complete official description of Task 4, all fourteen words of it:
Look at the same illustration from Task 3 and describe what you think will happen next.
Read the first six words again. The same illustration from Task 3.
Almost every lesson on this task teaches it on its own, as though a new picture appears. It does not. You have just spent sixty seconds looking at that picture and describing it out loud, and the notes you made for Task 3 are already your notes for Task 4.
You get 30 seconds to prepare and 60 seconds to speak, the same as Task 3.
CELPIP's advice for Task 3 comes down to five points. For Task 4 it is three, and in our words they are:
- Imagine, but stay realistic: base each prediction on what you can see people doing now.
- Ask what each person will do next, and how they will feel about it.
- Use the people and objects around them for ideas.
Strategy 2 is really two questions, and most people only answer the first one. Actions, and feelings.
🎥 The same picture answered twice, and one number changes (10:41)
Two 60-second answers on the same free practice question, with something to count while you listen. The reveal at 4:19 is what the whole lesson turns on.
Jump to: 0:31 Task 3 gets five strategies, Task 4 gets three · 1:53 your Task 3 notes are your Task 4 notes · 4:19 the reveal, five times and zero times · 6:03 one person, not the people · 7:48 what they will feel · 8:42 best sentence first, not last
Your thirty seconds of preparation are not starting from nothing.
In Task 3 you named things. A woman at the counter. A boy holding a red balloon. Two people looking at a menu board. Those names are the raw material for every prediction you are about to make, and you already said them out loud.
So do not describe the picture again. Task 3 was what is happening. Task 4 is what happens next. Repeating the description spends seconds on a task you already completed.
In your thirty seconds: look at the three or four things you named in Task 3 and write what each one does next. That is the whole plan.
The difference between a weaker and a stronger Task 4 answer usually comes down to one habit.
A weaker answer predicts about all the people, and it repeats itself because every prediction sounds the same. A stronger answer predicts about people it can point to, so the listener knows exactly who it means.
| Predicting about everybody | Predicting about someone |
|---|
| All the people will buy tickets | The man in the grey jacket will get to the counter and buy two tickets |
| Everyone is going to enjoy the show | The little girl with the balloon will not sit still through it |
| The people will eat their food | The couple near the window will finally get the table they have been waiting for |
The right-hand column is not longer by much. It is more specific, and specificity is what got credited.
And you already have those people. You named them in Task 3.
Strategy 2 asks two things, and this is the half almost nobody teaches:
What will each person do next, and how will they feel about it?
An action alone is a fact. An action plus a feeling is a prediction with a person inside it.
The man will finally reach the counter.
The man will finally reach the counter, and he will be relieved, because he has been standing in that line since the picture was taken.
Words worth having ready: relieved, disappointed, proud, embarrassed, excited, frustrated, surprised, satisfied.
Be careful about which feelings you claim. These are predictions, so hedge them. He will probably feel. She is going to be. They will most likely be.
Here is the idea that changes where you put your effort.
In a sixty-second task, your opening is the one sentence the listener is sure to hear in full, and it sets up everything after it. Your closing is the sentence most likely to be rushed or cut off by the clock.
So in a sixty-second task, put your strongest prediction first, not last.
That is the opposite of what most templates teach, and it is the opposite of what the old version of this page taught, which spent its whole fourth step on writing a conclusion.
What to do instead. Put your strongest, most specific prediction in your first sentence, right after you say what you are looking at. Do not save it. If the clock stops you before you finish the last thought, you have still said the best thing you had.
A closing sentence is fine if it arrives. It is not what you should be protecting.
For predicting
will · is going to · will probably · is likely to · I expect · I imagine · in a minute or two · any second now
For hedging
it looks like · it seems like · he will probably · she is most likely going to · I would guess
For connecting a prediction to what you can see
because · since · so · judging by · you can tell from
Say this: Judging by the empty tray in her hand, she is going to head straight for the counter, and she will probably be annoyed if the line is still that long.
Avoid I assume that. It appears in the Level 8 answer's docked line, and it tends to arrive attached to a prediction about everybody.
The picture, from Task 3. A busy cinema lobby. A long line at the ticket counter. A man in a grey jacket near the front of it. A child holding a balloon. Two staff behind the counter, one of them on the phone. A popcorn stand with nobody serving it.
What you already named in Task 3: the line, the man in grey, the child with the balloon, the staff member on the phone, the empty popcorn stand.
The answer.
Looking at the same picture, I think the man in the grey jacket is about to reach the front of the line, and he is going to be relieved, because he has clearly been standing there a while.
Once he gets his tickets, I imagine the line will start moving properly, especially if the woman on the phone finishes her call and opens the second till. Her colleague will be a lot less stressed once that happens.
The little girl with the balloon will probably not wait quietly much longer. She is already leaning away from her father, so I expect she will start pulling him towards the popcorn stand any minute now, and he is going to be embarrassed about it.
And I think somebody will have to staff that popcorn stand soon, because with a line this size, they are going to lose sales if nobody is there.
Notice what it does. Every prediction is about a named person or thing from the picture. Three of the four include a feeling. It opens with the strongest one. And nothing in it describes the picture, because that was Task 3.
Write the people and things you already named, and one prediction beside each.
man grey jacket → reaches counter, relieved
staff on phone → opens 2nd till, colleague less stressed
girl + balloon → pulls dad to popcorn stand
popcorn stand → someone has to staff it
Four lines, nouns and arrows. Then choose which one is your strongest and start with it.
Describing the picture again. That was Task 3. Every second you spend on it is a second not spent predicting.
Predicting about everybody. All the people, everyone, the crowd. Name someone.
Actions with no feelings. Half of strategy 2 is about how the people will feel. It is free marks and almost nobody takes them.
Saving your best line for the end. The rated evidence points the other way.
Predicting something the picture cannot support. Strategy 1 says be realistic and tie predictions to what you can see people doing. The man will win the lottery tonight is imagination without a picture behind it.
1. Name the people (3 min). Open any Task 3 picture and write the four things you would name. Then close it and write one prediction for each, from memory. That is the real skill, because on test day the picture is in front of you but your Task 3 answer is already gone.
2. Add the feeling (5 min). Take five plain predictions and add a feeling and a reason to each. She will leave early becomes she will probably leave early, and she is going to feel guilty about it, because the presentation is not finished.
3. Best first (5 min). Write four predictions, then reorder them so the most specific one is first. Record it that way and listen to how much stronger the opening ten seconds sound.
Related: Task 3: Describing a Scene, which is where your Task 4 material comes from · Fast Planning for Tasks · Scoring and Performance Standards.