Correspondents for the Associated Press (AP) compiled what was intended to be a damning statistical analysis demonstrating Israel’s disregard for civilian casualties when conducting air strikes that target the homes of militants in Gaza. But instead, AP inadvertently revealed the depth of its bias against Israel; its analysis utilizes a flawed methodology and is derived from a statistical sample missing essential data.
The report, “Israel house strikes kill mostly civilians” was published on Feb. 13, 2015. It analyses 844 Palestinians it claims were killed in 247 Israeli air strikes during the summer 2014 war with Hamas in Gaza. The analysis contends that “60 percent” of the fatalities were children, women or elderly men and that “among those killed were 96 confirmed or suspected militants – just over 11 percent of the total.” The authors, however, divulge that they could not categorize 240 dead males between the ages of 16 and 59 because their names “did not appear on AP searches of militant websites or on street posters honoring fighters.”
Incomplete Data Sample
The failure to identify 30 percent of the sample -which consisted of the crucial demographic in which militants are found – should have raised a red flag. If the authors lacked the statistical background to recognize how flawed the sample was, the editors should have understood that without this information, the analysis was unsuitable for publication. As it stands, it is entirely possible that nearly 40 percent of the fatalities were militants, rather than the 11 percent claimed.
Why is that difference so important? As investigative journalist Richard Behar pointed out in his recent critique of the AP story, studies of conflicts since World War II have found that as many as 85 to 90 percent of fatalities were civilians. Placed in the context of modern urban asymmetric warfare, Israel would come out favorably compared to others. We don’t know for sure because of the gaps in the sample data.
Worse still, there is evidence that the authors never intended the study to be anything like an objective analysis. The message reverberating throughout the article is Israeli indifference to Palestinian civilian casualties. The authors cite “U.N. figures” enumerating “at least 1,483 Palestinian civilians were killed in the war — 66 percent of the overall death toll of 2,205.” They disguise the true source of these figures – the Gazan Health Ministry – an entity controlled by Hamas. But worse than presenting unchallenged the figures provided by a entity controlled by a terrorist organization, the Associated Press fails to provide competing figures from a painstaking study carried out by a private research institute in Israel, the Meir Amit Intelligence and Information Center.
There is reason to suspect that AP’s omission of Israeli figures was intentional and not an oversight. The article indicates that the authors conferred with the Meir Amit Center. The authors, in fact, do cite the Center’s information when it serves their purposes. They note that the Center confirmed it is reasonable to assume that all women fatalities are civilians, a point favoring the Palestinian claims. But they ignore the main finding of the Meir Amit Center’s report on the Gaza conflict. The Center conducted an exhaustive study to identify the status of each Palestinian fatality that occurred during the Israeli military operation. Their study was able to confirm the status of 1598 fatalities, tallying 55 percent as militants and 45 percent as non-combatants. Yet, this information is omitted from the AP report. Even more incriminating, an earlier unpublished version of the AP report did cite the figure of 890 militants provided by the Meir Amit Center, but this information was removed for the published version of the AP report; so there is no question that AP was aware of this report.
Richard Behar published a probing analysis of the AP story, identifying many of the same issues discussed above. Behar, however, raised important questions about the methodology used by AP. He asked
So what was AP’s methodology for its current “examination” of the 247 airstrikes on houses? We’ll never know, because the wire service doesn’t tell us. What specific problems did they encounter that might have skewed or affected their research? We’ll never know, because AP doesn’t tell us those anecdotes.