Sex (or controversy) sells. Couple it with Pakistan, and you have got a big story.
Pakistan was caught being naughty again. In what seems to have become a regular occurrence, several news sites, global and local, reported that our nation tops the list of “most porn-searching countries”.
The headline was formed in a way that was meant to assert that it was Google that unveiled this information.
For those who learnt of the story from local news sites, you were probably told that the report was picked from Salon, a US-based current affairs website.
If you read Salon’s version, you will learn that it was picked from another liberal activist blog, Alternet.
In her article on Alternet, Carrie Weisman writes:
However, there was no link to the "data released by Google" which Carrie refers to.
What Google says about the report
Anyone who works in digital content production – blogger, journalist or editor – knows that Google seldom releases data on any matter whatsoever. Rather, its search data has always been available under many of its web analytics products, with Google Analytics and Google Trends being the most widely used.
So, where did Weismann get this data from?
My best guess would be Google Trends. But why take my word for it?
Here is what the relevant Google employee said in his email over this specific issue:
"Google has not released any data related to adult searches in Pakistan, so in the case of the referenced Salon.com article, there was no official source link.
"This could be a case of our publicly available Google Trends tool being used incorrectly. Any user can type in a search term and get a rough idea which countries are searching for a term the most. For example, what I've done in this Google Trends example shows that Pakistan is the top country in the world searching for 'cricket'. You can replace 'cricket' with other words and the results would vary.
"The bottom line is, Trends findings are indicative, but certainly not definitive. It doesn't reflect the content that is out there, just what people are searching for. Plenty of other variables should be considered: e.g. number of Internet users per country, gender spread, local colloquial terms."
There you go.
To clarify further
Google Trends is a nifty tool that illustrates search term popularity over a period of time. It can also narrow popularity by region. It has been put to good use, like in the US, where it tries to make flu season predictions.
It has also been used to spread misinformation in matters like porn habits across the world.
All the news stories on this porn popularity matter bank on the notion that Google Trends produces objective studies of real-time scenarios, which is simply not true.
The data from Google Trends gets normalised. This means that the results you see are achieved by dividing the numbers of search results of the term that you input with “a common variable, like total searches.”
Therefore, the results on Google Trends are relative, not absolute (or as Google puts it, 'indicative', not 'definitive').
Furthermore, the effectiveness of Google Trends in real-time scenarios is disputed at best.
In 2011, a study by Wellesley College on the predictability of US Congressional Elections through Google search trends concluded that there was no strong correlation between search popularity and likelihood of winning.
Just so everyone's clear on this, Carrie Weisman's 'information' could not have been derived from Google Analytics either.
Thing is, under Google Analytics, any user had the option of keyword (or search term) research, but once Google started securing its traffic with HTTPS in 2011, the availability of keyword data significantly reduced. Finally in 2013, the plug was pulled on this feature.
Therefore, the Alternet article is not based on data from Google Analytics. (Some would argue that Google Keyword Tool is a replacement; however, the Keyword Tool provides ideas on keyword quality which is more beneficial towards its advertisement product.)
The many caveats of Google Trends
The best way to find out the answer to this whole porn uproar is to open up Google Trends yourself, and go on a trends-search rampage for all possible porn-related words. If Pakistan features highly on most terms, you'll know it's popular.
But never will you know for sure, even for one single term, whether the search volume was highest in Pakistan compared to other countries.
Say you search for the term 'porn'. The following graph comes up:
|Screenshot of a Google Trends search for 'porn' in Pakistan.|
Notice that the trend for this keyword peaked in June 2012, and has been falling after that. So if Google Trends is the yardstick, then it seems Pakistanis are losing interest in searching for porn – a winning headline right there!
Similar observations are seen on the search for 'sex'. Apparently, Pakistanis were most interested in sex in November 2009, and have been getting bored with it ever since. Another fabulous ' authentic finding'.
|Screenshot of a Google Trends search for 'sex' in Pakistan.|
At this point, please do remember the Google representative said: "Plenty of other variables should be considered: e.g. number of Internet users per country, gender spread, local colloquial terms."
Do birds have sex?
Here is another fun factor: searching for trends with and without quotation marks (because these are two vastly different searches).
For instance, here's how Pakistan fares in a bird sex (sans quotation marks) search on Trends:
And here's how it fares in a "bird sex" (with quotation marks) search on Trends:
Searching without quotation marks includes every search hit which had the words 'bird' and 'sex' in it, including the question: 'Do birds have sex?'.
Searching with quotation marks reduces the results to Google hits on the exact term "bird sex", in which India leads and search volume in most of the world is too low to be tabulated.
See how it works? Google Trends is biased towards search term density. Now imagine the headline:
Indians have strange fascination with "bird sex"
But that's crazy. The term "bird sex" may not necessarily be that popular in India; it may actually have gotten more hits in the US. But if it's more popular in India than other terms are, India will be coloured darker.
With that reasoning, it's understandable why conservative Muslim countries like Pakistan are ranking so highly on these lists – for people in third-world countries like ours, the internet has long been the go-to avenue for all things taboo, mostly porn.
Using the internet for mostly porn and using the internet for porn along with other stuff are two different things.
Since Pakistan is guilty of the former, it ranks fairly high on most pornographic Google Trends. But if it maintains its porn search volume at the same level while doubling its search volume for other terms, it is bound to fall in these rankings.
Of responsible journalism
“We do our best to provide accurate data and to provide insights into broad search patterns, but the results for a given query, such as those reported in this story from Pakistan, may contain inaccuracies because the sample size is too small for the results to be statistically sound.”
The lesson for journalists? There is nothing wrong with using Google Trends, but using its data erroneously to build gross misinterpretation is not on.
Let us be honest: we know Pakistanis like their guilty pleasures. We know that from Google Trends. But equally true is the fact that we cannot prove it from Trends. And we definitely cannot say we rank highest in the world.
Also read: I deny watching porn publically, but …
Interestingly, the versions on Salon and Alternet cannot be accused of a false insinuation; they are not claiming anything beyond a general trend in porn interest. The versions in our local papers, though, are a different story.
It is best not to buy into this sensationalism. In this era of mass information, there always will be misinformation. There will always be articles with clickbait headlines and stories banking on controversy for their share value.
The onus lies upon us, the readers as much as the journalists.
We need to take a stand and share with responsibility.