Comparing a Graph with a Statement and Responding (Grafik und Aussage abgleichen, Stellung nehmen)
In the fourth task of the speaking part, you are in a seminar discussion. The lecturer shows a survey graph, a fellow student interprets it, and you respond: you check the statement against the numbers and then take a position (German: Stellung nehmen) yourself. The speaking time is 1:30 minutes.
The task
First you see the situation and a bar chart with the results of a survey, for about 30 seconds. Then you hear the statement of a simulated person through the headset, for about 20 seconds. This is followed by about 1:30 minutes of preparation and 1:30 minutes of speaking time (German: Sprechzeit), each with a countdown. Notes on paper are allowed.
Expect the statement to have two parts: a reading of the graph („Die Zahlen zeigen, dass …“) and a conclusion drawn from it, often a demand („Deshalb sollte die Hochschule …“). Your answer must take up both. The graph is your evidence: if you only give your opinion, you have left out half of the task.
The recording is assessed later by trained raters according to the provider's five criteria: relevance to the topic, task fulfilment, use of the sources, fluency and pronunciation, and language use. The result of the part is converted to 0 to 20 (5–9 TDN 3, 10–15 TDN 4, 16–20 TDN 5); raw-score cut-offs are not published. There are two sources here: the graph and the statement you heard.
| Criterion | How you gain | How you lose |
|---|---|---|
| Relevance to the topic | You stay with the survey question and the statement. | You talk about the topic in general, without the graph and the statement. |
| Task fulfilment | The comparison and your own justified position are both there. | You only describe the graph or only give your opinion. |
| Use of the sources | You name concrete values and report the statement correctly. | „Viele“ and „wenige“ instead of numbers, or you twist the statement. |
| Fluency and pronunciation | Numbers come out confidently, your thoughts are linked. | You falter at „einundvierzig Prozent“ or just read values off. |
| Language use | Phrases for graphs and polite disagreement, Subjunctive II. | „Das ist falsch“ as your only way of disagreeing. |
Time plan
At 120 to 140 words per minute, ninety seconds correspond to about 180 to 210 words. Because the preparation here is as long as the speaking time, you can plan thoroughly.
| Phase | Duration | What you do |
|---|---|---|
| Read the graph | about 0:30 | Survey question, who was asked, unit and footnote (multiple answers?), ranking, highest and lowest value. |
| Hear the statement | about 0:20 | Note the reading and the conclusion separately, note signal words like fast alle. |
| Preparation | about 1:30 | Comparison: which values test the reading? Fix your position and reason. |
| Speaking, reference | about 10 seconds | Name the statement and the graph. |
| Speaking, values | about 25 seconds | Two or three concrete values that matter for the comparison. |
| Speaking, comparison | about 30 seconds | Does the graph support the reading, partly or not at all? |
| Speaking, position | about 25 seconds | Your own stance on the conclusion with a reason, ideally with a counter-proposal. |
The note template for the preparation:
Aussage: Lesart … / Folgerung … Grafik: Wert … / Wert … / Wert … Abgleich: stützt / teilweise / widerspricht, weil … Ich: … weil … (Vorschlag: …)
While listening, there is no time for full sentences. Note two lines, one for the reading and one for the conclusion, and underline quantity words such as fast alle, die meisten, kaum jemand: that is usually exactly where the error you can expose lies.
When you run out of things to say. Say what the graph does not show: reasons, other groups and, if it covers only one year, developments over time. „Die Umfrage sagt allerdings nichts darüber, warum …“ is a sentence with content that deepens the comparison. Or develop your counter-proposal with an example.
When time gets short. Your position must not be missing. At about 20 seconds remaining, jump to „Deshalb halte ich … für …“, even if you still wanted to name a value. Two values and a position fulfil the task; five values without a position do not.
Step by step
- Take in the graph in 30 seconds. Read the title and the question first, then the ranking: what is in front, what is behind, where are the big gaps? Do not memorise every value, but the structure, and note the two or three most striking values: do not take it for granted that the graph stays visible during the preparation. If the graph shows values from two years, the most striking change is one of these values.
- Break the statement down. While listening, separate the reading of the data from the conclusion. The reading can be checked against the graph; the conclusion can only be evaluated. If you mix the two, you end up contradicting the wrong half.
- Check the reading value by value. Look for the two or three values that decide the reading. Typical distortions are exaggeration (fast alle for 27 percent), a swapped ranking, a relative majority passed off as an absolute one, a causal conclusion that a survey cannot support, and a reading that goes beyond the limits of the data (see the next section).
- Reach a verdict. The graph supports the statement, partly supports it or contradicts it. "Partly" is often the most accurate answer, and it shows that you differentiate. State the verdict explicitly, not just between the lines.
- Name values concretely. „Das Fahrrad kommt mit 27 Prozent erst an zweiter Stelle“ is evidence; „ziemlich viele fahren Rad“ is a claim. Every statement you make about the graph needs a number or a comparison.
- Take a position on the conclusion. Does the demand follow from the data, even if the reading was off? Often part of it is justified. Give reasons for your stance and, if possible, propose something better: a counter-proposal turns disagreement into a position.
- Stay polite but clear. You are speaking in a seminar. Refer to your fellow student by name and in the third person („Lena hat eben gesagt, dass …“) or speak of „die These“; that avoids the question of du or Sie. Disagreement begins with agreeing to the justified part.
Your turn1 of 5
According to a graph, 38 % of respondents eat lunch in the cafeteria, 30 % bring food, 22 % eat nothing and 10 % buy something on the way. A fellow student says: „Die Mehrheit isst in der Mensa.“ Which answer compares the statement with the graph best?
What the graph shows and what it does not
A reading can be off even though every number it mentions is right: it reads more into the graph than was measured. So check the labels during the 30 seconds of reading time: the question, the respondents, the unit and the footnote. Each of the nine limits below gives you a sentence with content for the comparison.
What exactly was asked
The question determines what a number means. If a survey asks which source students use most often for term papers and 15 percent name the library, it does not follow that only 15 percent use the library: the others may use it as a second source; they simply were not asked about that. In the same way, at least once a semester says nothing about how often, and how many respondents name an expense says nothing about how much money goes there.
Use is not demand
How many people use something shows what is offered and available, not what is wanted. If only 6 percent of students take the bus home at night, that may mean few need it, but just as well that it runs only once an hour or does not serve their neighbourhoods at all. A small share of users therefore does not prove small demand. Say in your comparison what is missing: a question about who would use the service if there were enough of it.
Who or what was counted
The graph applies only to the group that was asked and only to the unit that was counted. If only people in work were asked, everyone who does not work is missing; if a statistic covers only new buildings, it says nothing about the existing stock. If 40 percent of the households asked have no car, it does not necessarily follow that 40 percent of people live without a car, because a single-person household counts as much as a household of five. And if a survey asks the residents of a city where they go on holiday, it tells you nothing about the visitors who come to that city from elsewhere.
Multiple answers (Mehrfachnennungen)
If the graph says Mehrfachnennungen möglich (multiple answers possible), each person could choose several answers. The bars can then add up to more than 100 percent, and they do not exclude each other: someone who names „Ruhe“ (quiet) as a reason for the library may also have named „WLAN“ (Wi-Fi). So you may not add two bars („zusammen 90 Prozent“). You may compare the mentions: „Ruhe wird häufiger genannt als WLAN“ (quiet is named more often than Wi-Fi) is correct. Which reason matters more to an individual person, however, the graph does not show.
Without multiple answers, for instance on a scale like immer, meistens, selten, nie (always, mostly, rarely, never), the answers exclude each other. Then you may add neighbouring levels: 30 percent immer and 25 percent meistens make 55 percent who do it most of the time.
Two data points are not a trend
If the graph shows two years, you know the start and the end, but not the path in between and not the future. Whether the rise was steady, whether it is already levelling off or continuing, cannot be read from two values. If someone claims the development will continue, work out what that would mean. If a share rose from 20 to 36 percent in eight years, that is 2 percentage points a year, so at the same pace everyone would be reached only after another 32 years. „In ein paar Jahren alle“ (everyone in a few years) then does not follow even from the pace so far.
One year, several age groups: no development
If the graph compares age groups in a single year, it shows differences between people, not a change over time. If 18 percent of under-30s read a printed newspaper but 58 percent of over-60s do, it does not follow that the newspaper is dying out. Perhaps the younger people will themselves pick up a newspaper more often as they get older, perhaps they will keep their habit. Only data over several years can show which. A prediction like „In zwanzig Jahren liest niemand mehr Zeitung“ (in twenty years nobody will read a newspaper) cannot be checked against such a graph; say exactly that.
Between the groups or within one group
A graph with two groups can be read in two directions: across, by comparing the same value in both groups, and down, by looking at the ranking within one group. The two readings can diverge. If 35 percent of first-year students consider a sports programme important but only 22 percent of students in later semesters do, it matters more to the first-years than to the others. But if every other item comes ahead of it among the first-years, for example study spaces (78 percent) and Wi-Fi (70 percent), the sports programme is the least important item on the list for them too. „Den Erstsemestern ist das Sportangebot wichtiger als alles andere“ (the sports programme matters more to first-years than anything else) claims a ranking within the group, and the lead over the older students does not prove it. For every statement about a group, check which of the two readings it means, and give the matching figure.
Relative and absolute
With small starting values, relative changes sound enormous. If a share rises from 2 to 6 percent, it has tripled but is still only 6 percent. If the largest group stands at 50 percent, the gap of 44 percentage points is still large. In such cases name both, the growth and the level: „Das ist zwar eine Verdreifachung, aber auf sehr niedrigem Niveau.“ (That is a tripling, admittedly, but at a very low level.)
Index values
Some graphs show not shares but an index. A base year is set equal to 100 (2010 = 100), and every other value gives the ratio to that year. An index value of 130 means 30 percent more than in the base year, a value of 65 means 35 percent less, that is, a good third less. An index says nothing about the quantity itself: two curves that both start at 100 can stand for very different quantities.
Your turn2 of 5
A graph carries the note „Mehrfachnennungen möglich“. As their reason for going to the library, 58 % name „ruhiger Arbeitsplatz“ (a quiet place to work) and 47 % „schnelles WLAN“ (fast Wi-Fi). A fellow student says: „Den meisten ist Ruhe wichtiger als WLAN.“ (For most people quiet matters more than Wi-Fi.) Which answer compares this correctly with the graph?
Your turn3 of 5
A graph shows a rent index for a university town with „2015 = 100“. For 2025 it gives the value 140. Which sentence describes this correctly?
Typical pitfalls
Only describing the graph. For ninety seconds every bar is read out, and the fellow student's statement does not appear. A description is only the material for the comparison, not the answer.
Only opinion. „Ich finde auch, dass man mehr für Radfahrer tun sollte.“ Without a single value the graph remains unused, and use of the sources is missing.
Quantity words instead of numbers. „Viele fahren Bus, einige Rad, wenige Auto.“ That may be true, but it proves nothing. With „41 Prozent“ and „12 Prozent“ it becomes an argument.
All values, no selection. If you name every bar, you lose the time for the comparison and the position. Name only the values that decide the statement.
The statement reported wrongly. The fellow student demands that the car park be made smaller, the answer contradicts a demand to close it. If you refute a statement that was never made in that form, you miss the source.
Impolite disagreement. „Das ist doch Quatsch.“ In a seminar this is inappropriate, and it shows no range. „Da würde ich widersprechen“ is just as clear and suits the situation.
Causes from survey values. „Die Leute fahren Bus, weil das Semesterticket so günstig ist.“ The graph shows shares, not reasons. If you suspect a reason, mark it as an assumption.
Percent and percentage points mixed up. The gap between 41 and 27 percent is 14 percentage points, not 14 percent. If you confuse the unit, you make exactly the mistake you accuse your fellow student of.
Useful phrases
These useful phrases (German: Redemittel) suit a discussion in a seminar: factual, polite, spoken. Learn two from each group.
Referring to the statement
| Redemittel | Meaning and use |
|---|---|
| Ich möchte gern auf das eingehen, was Lena eben gesagt hat. | 'I'd like to respond to what Lena just said'; opening |
| Lena hat ja die These aufgestellt, dass … | 'Lena put forward the thesis that …'; reporting the reading |
| Daraus folgert sie, dass … | 'From this she concludes that …'; reporting the conclusion |
Referring to the graph and naming values
| Redemittel | Meaning and use |
|---|---|
| Wenn man sich die Zahlen genauer ansieht, … | 'If you look at the numbers more closely, …'; leading over to the graph |
| … liegt mit … Prozent klar vorn. | '… is clearly ahead with … percent'; highest value |
| … kommt mit … Prozent erst an zweiter Stelle. | '… only comes second with … percent'; ranking |
| Nur … Prozent geben an, dass … | 'Only … percent say that …'; low value, stressed |
| Das ist gut ein Viertel / knapp die Hälfte. | 'That is a good quarter / just under half'; putting a value into words |
| Der Abstand beträgt immerhin … Prozentpunkte. | 'The gap is as much as … percentage points'; difference |
| Der Anteil ist von … auf … Prozent gestiegen/gesunken. | 'The share has risen/fallen from … to … percent'; change between two years |
| … hat sich fast verdoppelt. | '… has almost doubled'; putting a change into words |
Comparing
| Redemittel | Meaning and use |
|---|---|
| Das wird von der Grafik gestützt/bestätigt. | 'The graph supports/confirms that'; statement is right |
| Das stimmt nur zum Teil. | 'That is only partly true'; statement partly right |
| Die Zahlen sprechen eher dagegen. | 'The numbers rather speak against it'; statement refuted |
| Von „fast allen“ kann man hier nicht sprechen. | 'You can't speak of "almost everyone" here'; exposing an exaggeration |
| Die Umfrage sagt allerdings nichts darüber, warum … | 'The survey says nothing, however, about why …'; limit of the data |
| Die Frage lautete ja nur, ob … | 'After all, the question was only whether …'; limit: what was asked |
| Befragt wurden allerdings nur … | 'However, only … were asked'; limit: who was asked |
| Da man mehrere Antworten geben konnte, darf man die Werte nicht addieren. | 'Since several answers were possible, you cannot add the values'; multiple answers |
| Aus zwei Messjahren lässt sich kein Trend ablesen. | 'Two years of data show no trend'; only two points in time |
| Das sind Unterschiede zwischen den Altersgruppen, keine Entwicklung über die Zeit. | 'These are differences between the age groups, not a development over time'; only one year, several groups |
| In der Rangfolge der Jüngeren selbst kommt … erst an letzter Stelle. | 'In the younger group's own ranking, … comes only last'; ranking within one group instead of comparing two groups |
| Das ist zwar eine Verdreifachung, aber auf niedrigem Niveau. | 'That is a tripling, admittedly, but at a low level'; small starting value |
Agreeing and disagreeing politely
| Redemittel | Meaning and use |
|---|---|
| Da hast du/hat Lena völlig recht. | 'You're/Lena is completely right there'; agreement |
| In einem Punkt stimme ich zu, nämlich … | 'I agree on one point, namely …'; partial agreement |
| Da würde ich widersprechen. | 'I would disagree there'; clear, polite disagreement |
| Das sehe ich etwas anders. | 'I see that a little differently'; soft disagreement |
| Ganz so einfach ist es, glaube ich, nicht. | 'I don't think it's quite that simple'; preparing an objection |
Taking a position and proposing
| Redemittel | Meaning and use |
|---|---|
| Deshalb halte ich es für übertrieben/sinnvoll, … | 'That's why I think it's exaggerated/sensible to …'; position |
| Ich fände es sinnvoller, … (with zu + infinitive) | 'I'd find it more sensible to …'; counter-proposal |
| Mein Vorschlag wäre also: … | 'So my suggestion would be: …'; close |
A worked example
Practice task, entirely invented. People, survey and numbers are made up.
Im Seminar „Verkehrsplanung“ zeigt Ihr Dozent, Herr Dr. Albers, Ergebnisse einer Befragung von 800 Studierenden Ihrer Hochschule und bittet die Gruppe, sie zu deuten. Als Erster meldet sich Ihr Kommilitone Tim zu Wort.
Wie kommen Sie überwiegend zur Hochschule? (800 Befragte)
Quelle: Erfundene Daten
Tim (heard): Also, die Grafik zeigt doch eindeutig, dass das Fahrrad inzwischen das Verkehrsmittel Nummer eins ist. Mit dem Auto kommt fast niemand mehr. Die Hochschule sollte deshalb den großen Parkplatz schließen und dort lieber überdachte Fahrradstellplätze bauen.
The statement broken down: reading 1 „Fahrrad Nummer eins“, reading 2 „fast niemand mit dem Auto“, conclusion „Parkplatz schließen, Fahrradstellplätze bauen“.
Notes during the preparation:
Aussage: Rad = Nr. 1? / Auto fast niemand? / → Parkplatz zu Grafik: Bus/Bahn 41 / Rad 27 (Platz 2) / Auto 12 (≈ 100 Leute) Abgleich: 1 widerspricht, 2 teilweise Ich: ganz schließen übertrieben, weil Gründe unbekannt; Vorschlag: Teil für Räder + Bus stärken
This turns into the following answer:
① Reference to the statement and the graph
Ich möchte gern auf Tims Beitrag eingehen. Die Umfrage zeigt ja, wie die Studierenden unserer Hochschule hauptsächlich zum Campus kommen.
② The deciding values
Wenn man sich die Zahlen genauer ansieht, liegen Bus und Bahn mit 41 Prozent klar vorn. Das Fahrrad kommt mit 27 Prozent erst an zweiter Stelle, danach folgen „zu Fuß“ mit 14 und das Auto mit 12 Prozent.
③ Comparison: reading 1 refuted, reading 2 partly right
Tims erste Aussage wird von der Grafik also nicht gestützt: Die Nummer eins ist nicht das Fahrrad, sondern der öffentliche Nahverkehr. Beim Auto hat er dagegen zum Teil recht, zwölf Prozent sind tatsächlich wenig. Aber von „fast niemand“ kann man nicht sprechen: Bei 800 Befragten sind das immerhin rund hundert Leute, und die Umfrage sagt nichts darüber, warum sie mit dem Auto kommen, zum Beispiel weil sie weit draußen wohnen.
④ Position on the conclusion, with a counter-proposal
Deshalb halte ich es für übertrieben, den ganzen Parkplatz zu schließen. Ich fände es sinnvoller, einen Teil davon für Fahrradstellplätze zu nutzen, denn immerhin gut ein Viertel kommt ja mit dem Rad, und gleichzeitig den Busverkehr zu stärken, weil der für die meisten am wichtigsten ist.
⑤ Close
Mein Vorschlag wäre also: mehr Platz für Fahrräder, aber nicht auf Kosten derer, die auf das Auto angewiesen sind.
193 words, about 1:30 minutes when spoken.
The answer names only four values, but exactly those that decide Tim's statement. It separates the two readings and judges each one on its own, agrees with the justified part and contradicts the rest with numbers. The position concerns Tim's conclusion, and the counter-proposal again rests on the graph.
Practise yourself
Fill in the phrases for working with the graph.
Your turn4 of 5
Your turn5 of 5
The preparation begins. What do you write on your sheet first?
Key takeaways
- Break the statement into reading and conclusion; you check the reading against the graph and evaluate the conclusion.
- Give an explicit verdict: supports, partly, contradicts.
- Name concrete values, but only those that decide.
- Watch for quantity words like fast alle and for causes that a survey does not show.
- Check the limits of the data: what was asked and who or what was counted, multiple answers, only two data points, one year with several age groups, comparison between groups or ranking within one group, small starting values, index values.
- Take a position on the conclusion and, where possible, make a counter-proposal.
- Disagree politely: first the justified part, then the objection.
- About 180 to 210 words fill 90 seconds; your position comes 20 seconds before the end at the latest.
Was this article helpful?